Using Information Technology and Social Networking for Recruitment of Research Participants: Experience From an Exploratory Study of Pediatric Klinefelter Syndrome

Download PDF
Monitoring Editor: Gunther Eysenbach
Reviewed by Malcolm Koo, Ibrahim Adeleke, and Shirley Williams
Sharron Close, MS, CPNP-PC, PhD,corresponding author1 Arlene Smaldone, CPNP-PC, DNSc,2 Ilene Fennoy, MPH, MD,3 Nancy Reame, MSN, PhD,2 and Margaret Grey, RN, DrPH1



Recruiting pediatric samples for research may be challenging due to parental mistrust of the research process, privacy concerns, and family time constraints. Recruitment of children with chronic and genetic conditions may further complicate the enrollment process.


In this paper, we describe the methodological challenges of recruiting children for research and provide an exemplar of how the use of information technology (IT) strategies with social networking may improve access to difficult-to-reach pediatric research participants.


We conducted a cross-sectional descriptive study of boys between the ages of 8 and 18 years with Klinefelter syndrome. This study presented unique challenges for recruitment of pediatric participants. These challenges are illustrated by the report of recruitment activities developed for the study. We reviewed the literature to explore the issues of recruiting children for research using conventional and IT approaches. Success rates of conventional recruitment approaches, such as brochures, flyers in medical offices, and physician referrals, are compared with IT-based outreach. The IT approaches included teleconferencing via a Klinefelter syndrome support group, services of a Web-based commercial recruitment-matching company, and the development of a university-affiliated research recruitment website with the use of paid advertising on a social networking website (Facebook).


Over a 3-month period, dissemination of over 150 recruitment brochures and flyers placed in a large urban hospital and hospital-affiliated clinical offices, with 850 letters to physicians and patients were not successful. Within the same period, face-to-face recruitment in the clinical setting yielded 4 (9%) participants. Using Web-based and social networking approaches, 39 (91%) agreed to participate in the study. With these approaches, 5 (12%) were recruited from the national Klinefelter syndrome advocacy group, 8 (19%) from local and teleconference support groups, 10 (23%) from a Web-based research recruitment program, and 16 (37%) from the university-affiliated recruitment website. For the initial 6 months, the university website was viewed approximately 2 to 3 times per day on average. An advertisement placed on a social networking site for 1 week increased website viewing to approximately 63 visits per day. Out of 112 families approached using all of these methods, 43 (38%) agreed to participate. Families who declined cited either travel distance to the study site (15, 22%) or unwillingness to disclose the Klinefelter syndrome diagnosis to their sons (54, 78%) as the reasons for nonparticipation.


Use of Web-based technologies enhances the recruitment of difficult-to-reach populations. Of the many approaches employed in this study, the university-affiliated recruitment website supported by a Facebook advertisement appeared to be the most successful. Research grant budgets should include expenses for website registration and maintenance fees as well as online advertisements on social networking websites. Tracking of recruitment referral sources may be helpful in planning future recruitment campaigns.

Keywords: patient recruitment, research subject recruitment, health information technology, social networking, Klinefelter syndrome


Recruiting children for research can present many challenges due to parental mistrust of the research process, privacy concerns, and family time constraints [1,2]. Children with chronic and genetic conditions may further complicate the recruitment process [3]. For studies conducted in the United States, additional challenges exist including regulations and guidelines that direct how researchers contact and enroll participants for studies.

Prior to 1996, medical treatment of children was based on clinical trials and the testing of products and medications that were conducted in adults [4]. Although many treatments were effective for adults, some were shown to be ineffective or harmful to children [4]. In 1996, members of a joint workshop of the American Academy of Pediatrics and the National Institute of Child Health and Development issued a consensus statement calling for children to receive adequately tested treatments recommending efforts to include children in research [4]. In1998, the National Institutes of Health (NIH) established a policy with guidelines requiring that children must be included in all human subjects research that is conducted or supported by the NIH unless there was a scientific or ethical reason for not doing so [5]. The goal of this policy was to increase child participation in research for the purpose of generating data specific to the treatment of children [4,5]. However, inadequate representation of children in selected areas of low-prevalence diseases, orphan conditions (such as Klinefelter syndrome), and genetic conditions persist [6]. More innovative approaches to the enrollment of pediatric volunteers in clinical studies are needed.

The general public increasingly uses information technology (IT) as a source of health information. Approximately 80% of the American public search for health information using Internet sources [7]. Researchers are now turning to the Internet as a tool to recruit target study populations [8-14] for Internet-based interventions in conditions such as hypertension [15], diabetes [16-18], smoking cessation [8,12,19], human immunodeficiency virus risk management [13], and depression [14].

The Internet offers many opportunities for informing potential research participants about a study. These include email [8,20,21], discussion boards, blogs [8], search engines [20,21], study websites [20-22], and Web-based platforms for matching researchers with participants. Each form of Web-based communication provides opportunities and challenges for subject recruitment.

On the one hand, exposure of information to vast numbers of Internet users creates an enormous opportunity for visibility and communication with potential research participants. At the same time, the recruitment process can be sabotaged by problems on the Internet, such as emails sent to spam folders [8-10], discussion board and blog administrators blocking content associated with the researcher [9], and poor choice or lack of adequate keywords on study websites that diminish search engine exposure [23]. A key limitation of the use of Internet-based recruitment activities include the inability to reach socioeconomically or educationally disadvantaged groups as well as culturally diverse populations who may lack access to the Internet or familiarity with its use [10]. In the same vein, potential recruits to a study may not be receptive to unsolicited emails, or may not trust the legitimacy of the sender [8]. Clinical investigators encounter special challenges when attempting to recruit children as research volunteers, especially those who have low-prevalence diseases or genetic conditions [6,24].

The purpose of this paper is to describe methodological challenges associated with the recruitment of children as volunteers in research and to discuss how IT may improve access and enrollment of children in research. A case study of our experience in recruiting boys with Klinefelter syndrome for an exploratory cross-sectional study is used to illustrate specific challenges encountered with a difficult-to-recruit pediatric research population and how IT was used to support the enrollment of participants in the study.

Research Recruitment Challenges

Successful enrollment of clinical research participants is both a science and an art. A number of factors, including patients, health care professionals and researchers, structural and organizational entities, and history, interact to compromise or undermine successful enrollment of patient volunteers into a clinical study [1]. Patients may have limited access to research information and might not fully understand the role of clinical research in the advancement of knowledge for drug and behavioral therapy development [1,21]. They may also worry about or mistrust researchers and their institutions due to lack of understanding about the research process or associated risks and benefits [1,21,25]. Patient characteristics, such as culture, language, and religion [1], may further reduce the chance of successful enrollment. Health care providers may play an important role in gaining access to potential participants, but also may represent barriers to such access [1,2,21].

Health care providers in nonacademic settings may have a limited understanding or interest in clinical trials or may have misgivings about academic institutions [1,26]. Community health care providers may also have concerns about losing control over their patient’s care, or losing the patient to another provider [1,26]. Full-time clinicians are frequently pressed for time in caring for patients and may be concerned about the additional administrative workload and lack of administrative support for research activities [1,2,26]. This concern may lead to financial disincentives for clinical providers to become involved in informing their patients about research enrollment opportunities [26]. Researchers themselves sometimes fail to recognize how they may contribute to recruitment and enrollment problems in their own studies. Lack of training and proficiency in communication for the conduct of research with low-literacy populations may lead to misunderstandings between the researcher and potential participant and, in turn, lower response rates of participation [1].

Other barriers facing researchers include lack of attention to the mistrust of the population to be recruited, failure to demonstrate cultural sensitivity, and lack of training in understanding health care disparities in underserved populations [1]. These barriers are often unrecognized by researchers and get in the way when attempting to gather the desired sample. Structural and organizational factors may also be associated with the desire or ability of people to volunteer for research [1]. Researchers need to consider logistic arrangements to facilitate patient participation, such as creating convenient times and locations for study participation.

Communities may also be sensitive about allowing researchers entry into their environment, especially when they perceive that their participation in the scientific efforts does not result in any return or reward at the community level [1]. This concern makes it very difficult for researchers to re-enter the same community or for the community to be approached by other researchers.

The history of disreputably negative research practices persists in the minds of the public and these perceptions may influence the attitudes of potential research volunteers. The awareness of inhumane treatment by Nazi researchers during World War II and the infamous Tuskegee Syphilis Project conducted by the US Public Health Service from 1932 to 1972 [27] may promote overall fear and mistrust about the research process in the minds of many potential research participants.

Levels of Protection that Challenge Research Recruitment

In the United States, several guidelines offer protection to the public with regard to personal health care and participation in research. Public Law 104-191, also known as the Health Insurance Portability and Accountability Act (HIPAA), was enacted to protect the privacy and personal health information of the public [28]. The HIPAA requirements also guide researchers on how to protect the privacy of research participants. Although all health care providers and researchers are required to obey these laws, many members of the public may be apprehensive of the attendant side effects of disseminating private medical information by researchers. Levels of protection, designed to benefit the public, also may impede progress in the timing and accomplishment of recruitment.

Although pediatric researchers are charged with the responsibility of recruiting children for research, several challenges exist in such efforts. Because parents are legally responsible for their children, it is the parent who must be approached for permission for their child to participate. Parents’ willingness to have their child participate in a study may be influenced by their perception of benefits, risks, and barriers to participation [2]. The child must also assent to the activities of the research project. The child’s willingness to participate in the project may depend upon his/her developmental status and any vulnerability related to illness, chronic condition, or communication disabilities. Children may view research participation as a positive experience, including a wish to help others, reward incentives, and the desire to have a fun experience [2]. These positive motivations may be offset by anticipated unpleasantries, such as blood tests, disagreeable medication regimens, or interruptions in their daily lives [29]. The child-recruit is embedded within a family with complex daily schedules often including parental work, school schedules, and sport practices or other extracurricular activities. All members of the family, including the child’s siblings, influence the busy family schedule. Researchers must anticipate and accommodate time commitments of the family as well as considerations for transportation and commute time. Finally, the parents and the child-recruit must be prepared to agree about certain participation risks and unpleasantries such as completing multiple forms and surveys, or medical examinations, including blood collection.


Case Illustration: A Study of Boys With Klinefelter Syndrome

The exemplar case illustrates our recent experience with recruiting boys with Klinefelter syndrome for participation in a cross-sectional study. Traditional approaches to recruitment fell short of obtaining the desired sample and expanding the approach with IT resulted in a significant gain in enrollment.

Klinefelter syndrome is a genetic condition caused by the presence of an extra X chromosome (karyotype 47, XXY). This condition occurs in an estimated 1 in 450-500 male births [30,31]. Although it is not rare, it is extremely underdiagnosed. Approximately 64% of affected males are not aware of the diagnosis, and of the 36% who are aware, only 10% are diagnosed in childhood [32]. Klinefelter syndrome in adults is associated with androgen deficiency, gynoid distribution of body fat, gynecomastia, small testes, and azoospermia [33,34]. Individuals diagnosed with Klinefelter syndrome during adulthood report childhood developmental delay; speech, language and learning problems; and psychological issues including depression, shyness, aggression, and social interaction difficulties [35,36]. Klinefelter syndrome poses increased health risks throughout the life span, including increased risk for cardiovascular disease, diabetes, and osteoporosis [37]. Diagnosis of Klinefelter syndrome during childhood may represent an opportunity to address both physical and psychosocial health challenges.

Klinefelter syndrome is a misunderstood condition owing to a paucity of research in children, lack of clear clinical guidelines for treatment during life stages, and unfortunate conclusion errors made by early researchers that suggested men with Klinefelter syndrome were at increased risk for criminal behavior [38-40]. As a result, Klinefelter syndrome families may struggle with inadequate information, lack of support, perceived stigma, and uncertainties about their son’s health [41]. Current research focused on boys with Klinefelter syndrome report fairly small sample sizes, ranging from groups of less than 20 [42-44] to the largest reported cohort of 93 [45]. Misunderstandings about Klinefelter syndrome may contribute to reluctance on the part of many men and families of young sons with Klinefelter syndrome to discuss or disclose information about their diagnosis to others [24].

We conducted an exploratory descriptive study to better understand phenotype, biomarkers, and psychosocial health parameters of boys with Klinefelter syndrome between the ages of 8 and 18 years [46]. The study protocol included a physical examination, blood collection for reproductive and cardiovascular biomarkers, and psychosocial health measurements including quality of life, self-esteem, self-concept, and risk for depression. The Columbia University Institutional Review Board approved the protocol for this study. For this exploratory study, sample size was based on a moderate correlation of at least 0.40 between the clinical characteristics and psychosocial variables as observed in studies of health-related quality of life and polycystic ovary syndrome [47,48]. For a correlation of 0.40 with alpha=.05, a total of 46 subjects were required for a minimum power of 80%. No previous studies with a Klinefelter syndrome population studied the relationship between clinical characteristics and psychosocial health. Recruitment was planned with traditional approaches, including contacting patients in a local pediatric endocrine practice; sending letters to pediatricians, pediatric endocrinologists, geneticists, and genetic counselors; and the use of recruitment flyers and brochures placed strategically throughout the medical center. After sending 850 letters, placing 150 brochures and fliers, and approaching 23 families during clinical visits, only 4 boys were recruited in a 3-month period. It became readily apparent that the traditional approach would fail to achieve the minimum sample size of 46 according to our sample size calculation. Thus, a more innovative approach was devised using IT and social networking.

New recruitment strategies included the development of a study website, in-person information sessions, Web-links, teleconferences, and email access to members of a national and several regional Klinefelter syndrome support organizations, as well as registration with a computer platform clinical recruitment-matching service. Each strategy is briefly described subsequently.

Klinefelter Syndrome Study Recruitment Website

A study information and recruitment website [49] was created using the keywords Klinefelter syndrome, KS, boys with KS, and KS phenotype to increase the likelihood that people searching the Internet for information on Klinefelter syndrome might find the website when conducting searches. The website pages provided information regarding the study and its eligibility requirements, study procedures, and how to contact the researcher for further information or enrollment. The study website home page screenshot is shown in Multimedia Appendix 1.

Patient Advocacy Associations

We contacted a national Klinefelter syndrome advocacy association, Knowledge Support & Action (KS&A) [50], who agreed to place information about our study with the study website link on their website. A screenshot of KS&A home page is provided in Multimedia Appendix 2. Regional Klinefelter syndrome support groups with links to the national organization then invited us to give live presentations about our study at their meetings and also agreed to send emails about the presentation and the study to their members. One of the regional groups, the Klinefelter Syndrome Global Support Group (screenshot is shown in Multimedia Appendix 3), offered a monthly parent teleconference. Over a 3-month period, we were able to explain the purpose of the study and to respond to questions regarding our protocol.

Web-Based Clinical Recruitment-Matching Service

RecruitSource is a search engine and computer platform for matching clinical research participants with researchers [51]. A screenshot of the RecruitSource home page can be seen in Multimedia Appendix 4. Researchers can register details about their study and provide eligibility requirements for matching with potential participants.

Patients who might be interested in research participation register their health information via PrivateAccess [52] as shown in Multimedia Appendix 5. This website is a secure Internet registry that enables them to control who can and cannot see all or selected parts of their personal health information. This IT-based platform prescreens the potential participants who give advance privacy directives about their health information and are asked whether they wish to be contacted by a researcher. The incentive for people using this registry is that they can share their personal health information with properly authenticated doctors, researchers, or family members on a secure Internet platform. All contact information is coded and encrypted for privacy. The potential participant gives specific permission to be contacted by the researcher. Once the patient is registered, the researcher receives information about participants who have expressed an interest in being contacted for possible inclusion in the study. This service is provided at no cost to the researcher if the RecruitSource Web link is accessed via a patient advocacy association. In this case, the study was linked to the KS&A organization, a national advocacy association for Klinefelter syndrome [50].

Social Networking

Social networking is often defined by Web-based platforms, such as Facebook and others. Social networking, however, may also include face-to-face and teleconference transactions with groups, audiences, researcher-participant, and participant-participant networking. Participant-participant networking is the central component to the recruitment strategy known as snowballing [53]. We used all these networking processes in our Klinefelter syndrome study. The interlinking of IT-based and face-to-face networking provided an opportunity for multiple modes of information exposure about the study. Midway into recruitment, we decided to conduct a short trial of a Facebook advertisement (ad) as shown by the screenshot in Multimedia Appendix 6. Because we had not anticipated this strategy a priori, funding for advertising was limited. Nevertheless, we wished to observe how a 1-week social networking ad might impact exposure to the study website.


Of 112 families approached, 43 (38%) agreed to participate. The most frequent reasons for families declining participation was nondisclosure of the diagnosis to their sons (54/112, 78%) and geographic distance from the study site (15/112, 22%). Most parents who had not disclosed the diagnosis to their sons feared that their sons would learn of the diagnosis through participation.

Recruitment approaches for the participants in the Klinefelter syndrome study are summarized in Table 1. Recruitment using IT and social networking yielded a greater number of participants (39/43, 91%) compared to use of traditional approaches (4/43, 9%).

Table 1

Number of participants using traditional and information technology with social networking recruitment approaches to the Klinefelter syndrome study (N=43).

Of the 69 families who declined, over one-fifth (15/69, 22%) came from direct clinical contact; almost twice that number (29/69, 42%) declined during support group presentations, and one-sixth (10/69, 16%) declined during the national KS&A meeting. The most frequent reason for decline was parents not wanting their boys to learn of their diagnosis (n=54, 78%) and travel distance to study site (15/69, 22%).

In an effort to boost activity from general Web users, we placed an ad on Facebook. The ad ran for 1 week in June of 2010, targeting a general audience. Impressions are the raw number of times an ad is shown to different Facebook users. The Facebook ad was shown a total of 2,522,169 times. Social impressions reflect the number of times the ad was shown with social context who visited the study webpage. There were 2835 social impressions for this ad resulting in 509 clicks directly to the study website’s home page. At a total cost of $311 for the week’s ad, this represents the researcher’s cost of $0.61 per visit. Prior to placing the ad on Facebook, the study website received 2 to 3 visits per day. During the week of Facebook advertising, website visits climbed to an average of 63 visits per day. The Klinefelter study website activity increase in response to the Facebook ad can be seen in Figure 1.

Figure 1

Response (visits per day to the Klinefelter syndrome study website) during Facebook advertisement period showing increase in activity during Facebook advertising.

Because multiple techniques were employed to attract this difficult-to-reach population, it is difficult to attribute any one recruitment approach to increasing the number of participants in this study. Figure 2 shows a timeline of the 1-year recruitment process.

Figure 2

Recruitment to the Klinefelter syndrome study by source timeline.


The Internet represents an increasingly valuable resource for researchers, especially for those who wish to understand the social and cultural context of the populations they are attempting to reach [54]. Since the advent of IT as a common mode of communication, researchers have learned many lessons about the pearls and pitfalls of using this recruitment approach. In difficult-to-reach and vulnerable populations, such as Klinefelter syndrome families, our experience has led us to understand better that more than 1 recruitment technique may be required to inform potential participants and to foster trust in them. We believe that the construction and launch of the study website served these 2 important purposes. We were, however, unable to attribute increased participation due to any 1 technique, including placement of the Facebook ad. Important lessons were learned during this challenging recruitment process, such as the need to track how participants make decisions about whether to participate. We were unable to track which potential research candidates came from the study website activity while the Facebook ad ran because we did not have access to the server log. Anecdotally, several families reported that they chose to participate only after being exposed to the study information from multiple sources. Some families reported that friends or other family members who saw the Facebook ad contacted them to let them know about the study. Once informed, these families conducted either a general Internet search, visited the website directly, or visited the KS&A website for more information. Most importantly, we discovered that we need to track sources of recruitment more carefully in the future by surveying participants about how they found out about the study and also by looking at server logs whenever possible. It would also be helpful to track website visits by Internet protocol (IP) addresses to examine how many potential candidates are first-time or repeat visitors. Because we were unable to attribute which of our recruitment responses came from the Facebook ad, we are unable to estimate the cost per participant. This information would have been very helpful in planning cost allocation for a future study.

Since the advent of IT and social networking in the scientific community, there has been a steady evolution of its use for recruitment and Internet-based interventions. Even within the past 5 to 7 years, much has been learned about the limitations of Internet-based approaches and how such problems might be mitigated.

Although early experience with the use of IT-based recruitment for clinical research, as reported by Koo and Skinner [8], was disappointing, others have offered solutions to optimize challenges that make this form of recruitment difficult. Murray et al [20] solved issues related to mass emailing and spam management by providing recipients with the option to unsubscribe in order to decline further contact by researchers. They were also able to demonstrate the benefits of advertising their study on the home page of a well-known and trusted charity. Our Klinefelter syndrome study recruitment was greatly enhanced by our exposure with the KS&A national advocacy association and with support groups. Ip et al [9] addressed IT recruitment challenges by developing a guide describing a 12-step process to improve visibility and popularity of recruitment messages. The goal of this guide was to increase the interest of potential participants and to offer researchers ways in which to anticipate and respond when IT communication difficulties arise. Recent work with Ramo and Prochaska [11] demonstrated the value of Facebook advertising as an effective mechanism to reach young adults in clinical research. However, reaching a target group under the age of 18 years imposes special issues. For example, although children can be attracted to recruitment advertising for research, they would still be required to obtain parental consent for participation. Although social networking may interest a child about a research project, additional means of informing and developing trust with a parent are still necessary. Sullivan et al [13] and Graham et al [12] each illustrated how changing the composition of banner advertising may improve communication to desired target groups. In the case of pediatric research, such customization may promote discourse between parents and children. The recruitment process, as described by Patel et al [21], is explained as a dialog or discourse that takes place between the investigator and the potential research participant. In the case of minor children, discourse needs to be promoted between investigator, parent, and child if pediatric recruitment is to be successful.

Our recent experience in recruiting boys for the Klinefelter syndrome study can be described as a multilayered strategy of communication using IT. The process of communication began with a traditional print exposure that proved to be ineffective. Adding the various IT communication approaches, including the study website, a computer-based research recruitment website, social networking on Facebook, exposure via support groups online, and by teleconference, offered parents multiple exposures to study information. Although our original sample size calculation called for 46 participants based upon an effect size of 0.40, the effect size from our study proved to be larger (–0.47). We believe that the overall number of participants (43 boys) did not negatively affect the study.


Recruiting boys for a study on Klinefelter syndrome proved to be a challenging endeavor that was best accomplished using IT-based techniques. Important lessons were learned as we dealt with early recruitment challenges. The first lesson is that multiple exposures to the study information and personal contact with the researcher may be helpful in fostering parental trust. Parents must believe that the study, the institution, and the researcher are trustworthy before they will agree to have their child take part. These acts of communication, presented in multiple ways, were central to the success of our recruitment effort and are distinct advantages offered by IT-based strategies. Expenses related to website creation, registration of a domain name, website maintenance, and planning for social networking advertising were not initially anticipated by us, but should be considered by future researchers in the planning process when study budgets are developed. A limitation of our reported recruitment observations is that an in-depth recruitment analysis was not conducted to determine how multiple recruitment exposures occurred. Future IT-based recruitment efforts should preplan the collection of profile data, including IP addresses and tracking of how, when, and how many times a recruitment website was visited. This type of data may assist in the planning of customized approaches for the creation of more effective social networking banner advertising. Nevertheless, the observations from this study may advance the understanding of how difficult-to-recruit participants, like children, might be reached and have parental communication needs met with a view to obtaining their consent to participate in a study. It is noteworthy to mention that there has been inadequate representation of children in Klinefelter syndrome research and in other genetic conditions.

Researchers need to expand their knowledge of how potential recruits might be encouraged to participate in studies by understanding the utility of traditional approaches versus IT and other social networking approaches for recruitment. By offering multiple opportunities for exposure, parents have the opportunity to digest and think about the idea of having their child participate in a study. Because IT and social networking have become well-accepted modes of communication, these tools enable the researcher to layer the recruitment message in order to optimize the likelihood that recruitment efforts will be successful.


This research was supported by The Pediatric Endocrinology Nursing Society, the National Association of Pediatric Nurse Practitioners, the Alpha Zeta Chapter of Sigma Theta Tau, Columbia University’s CTSA grant No. UL1 RR024156 from NCATS-NCRR/NIH and Yale School of Nursing T32 Post-Doctoral Fellowship Training Grant, No. 5T32NR008346-08. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.


Internet protocol
information technology
Klinefelter syndrome

Multimedia Appendix 1

Screenshot of Columbia University Klinefelter syndrome study website.

Multimedia Appendix 2

Screenshot of Knowledge Support & Action (KS&A) Klinefelter syndrome advocacy association website.

Multimedia Appendix 3

Screenshot of Klinefelter Syndrome Global Support Group website.

Multimedia Appendix 4

Screenshot of RecruitSource website.

Multimedia Appendix 5

Screenshot of PrivateAccess website.

Multimedia Appendix 6

Screenshot of Klinefelter syndrome study Facebook advertisement.


Conflicts of Interest:

Conflicts of Interest: None declared.


1. Baquet CR, Henderson K, Commiskey P, Morrow JN. Clinical trials: the art of enrollment. Semin Oncol Nurs. 2008 Nov;24(4):262–9. doi: 10.1016/j.soncn.2008.08.006. [PMC free article] [PubMed] [Cross Ref]
2. Caldwell PH, Murphy SB, Butow PN, Craig JC. Clinical trials in children. Lancet. 2004;364(9436):803–11. doi: 10.1016/S0140-6736(04)16942-0. [PubMed] [Cross Ref]
3. Gallo AM, Hadley EK, Angst DB, Knafl KA, Smith CA. Parents’ concerns about issues related to their children’s genetic conditions. J Spec Pediatr Nurs. 2008 Jan;13(1):4–14. doi: 10.1111/j.1744-6155.2008.00129.x. [PubMed] [Cross Ref]
4. National Institutes of Health. 1998. Mar 06, [2012-06-01]. webcite NIH policy and guidelines on the inclusion of children as participants in research involving human subjects
5. US Department of Health and Human Services, National Institutes of Health. 1999. Mar 17, [2012-06-01]. webcite Grants and funding: inclusion of children as participants in research involving human subjects
6. Martinez-Castaldi C, Silverstein M, Bauchner H. Child versus adult research: the gap in high-quality study design. Pediatrics. 2008 Jul;122(1):52–7. doi: 10.1542/peds.2007-2849. [PubMed] [Cross Ref]
7. Fox S, Fallows D. Health searches and email have become more commonplace, but there is room for improvement in searches and overall Internet access. Washington, DC: Pew Internet & American Life Project; 2003. Jul 16, [2012-06-01]. webcite
8. Koo M, Skinner H. Challenges of internet recruitment: a case study with disappointing results. J Med Internet Res. 2005;7(1):e6. doi: 10.2196/jmir.7.1.e6. [PMC free article] [PubMed] [Cross Ref]
9. Ip EJ, Barnett MJ, Tenerowicz MJ, Perry PJ. The touro 12-step: a systematic guide to optimizing survey research with online discussion boards. J Med Internet Res. 2010;12(2):e16. doi: 10.2196/jmir.1314. [PMC free article] [PubMed] [Cross Ref]
10. Suarez-Balcazar Y, Balcazar FE, Taylor-Ritzler T. Using the Internet to conduct research with culturally diverse populations: challenges and opportunities. Cultur Divers Ethnic Minor Psychol. 2009 Jan;15(1):96–104. doi: 10.1037/a0013179. [PubMed] [Cross Ref]
11. Ramo DE, Prochaska JJ. Broad reach and targeted recruitment using Facebook for an online survey of young adult substance use. J Med Internet Res. 2012;14(1):e28. doi: 10.2196/jmir.1878. [PMC free article] [PubMed] [Cross Ref]
12. Graham AL, Fang Y, Moreno JL, Streiff SL, Villegas J, Muñoz RF, Tercyak KP, Mandelblatt JS, Vallone DM. Online advertising to reach and recruit Latino smokers to an internet cessation program: impact and costs. J Med Internet Res. 2012;14(4):e116. doi: 10.2196/jmir.2162. [PMC free article] [PubMed] [Cross Ref]
13. Sullivan PS, Khosropour CM, Luisi N, Amsden M, Coggia T, Wingood GM, DiClemente RJ. Bias in online recruitment and retention of racial and ethnic minority men who have sex with men. J Med Internet Res. 2011;13(2):e38. doi: 10.2196/jmir.1797. [PMC free article] [PubMed] [Cross Ref]
14. Meyer B, Berger T, Caspar F, Beevers CG, Andersson G, Weiss M. Effectiveness of a novel integrative online treatment for depression (Deprexis): randomized controlled trial. J Med Internet Res. 2009;11(2):e15. doi: 10.2196/jmir.1151. [PMC free article] [PubMed] [Cross Ref]
15. Green BB, Anderson ML, Ralston JD, Catz S, Fishman PA, Cook AJ. Patient ability and willingness to participate in a web-based intervention to improve hypertension control. J Med Internet Res. 2011;13(1):e1. doi: 10.2196/jmir.1625. [PMC free article] [PubMed] [Cross Ref]
16. Glasgow RE, Strycker LA, Kurz D, Faber A, Bell H, Dickman JM, Halterman E, Estabrooks PA, Osuna D. Recruitment for an internet-based diabetes self-management program: scientific and ethical implications. Ann Behav Med. 2010 Aug;40(1):40–8. doi: 10.1007/s12160-010-9189-1. [PubMed] [Cross Ref]
17. Grey M, Whittemore R, Liberti L, Delamater A, Murphy K, Faulkner MS. A comparison of two internet programs for adolescents with type 1 diabetes: design and methods. Contemp Clin Trials. 2012 Jul;33(4):769–76. doi: 10.1016/j.cct.2012.03.012. [PMC free article] [PubMed] [Cross Ref]
18. Whittemore R, Jaser SS, Jeon S, Liberti L, Delamater A, Murphy K, Faulkner MS, Grey M. An internet coping skills training program for youth with type 1 diabetes: six-month outcomes. Nurs Res. 2012;61(6):395–404. doi: 10.1097/NNR.0b013e3182690a29. [PMC free article] [PubMed] [Cross Ref]
19. Sadasivam RS, Delaughter K, Crenshaw K, Sobko HJ, Williams JH, Coley HL, Ray MN, Ford DE, Allison JJ, Houston TK. Development of an interactive, Web-delivered system to increase provider-patient engagement in smoking cessation. J Med Internet Res. 2011;13(4):e87. doi: 10.2196/jmir.1721. [PMC free article] [PubMed] [Cross Ref]
20. Murray E, Khadjesari Z, White IR, Kalaitzaki E, Godfrey C, McCambridge J, Thompson SG, Wallace P. Methodological challenges in online trials. J Med Internet Res. 2009;11(2):e9. doi: 10.2196/jmir.1052. [PMC free article] [PubMed] [Cross Ref]
21. Patel MX, Doku V, Tennakoon L. Challenges in recruitment of research participants. Advances in Psychiatric Treatment. 2003;9:229–238. doi: 10.1192/apt.9.3.229. [Cross Ref]
22. Linke S, Murray E, Butler C, Wallace P. Internet-based interactive health intervention for the promotion of sensible drinking: patterns of use and potential impact on members of the general public. J Med Internet Res. 2007;9(2):e10. doi: 10.2196/jmir.9.2.e10. [PMC free article] [PubMed] [Cross Ref]
23. Lowery S. Web-Source. 2012. [2012-06-01]. webcite Website optimization
24. Herlihy AS, Halliday JL, Gillam LH. Ethical issues in recruiting prenatally diagnosed adults for research: Klinefelter syndrome as an example. Public Health Genomics. 2012;15(1):31–3. doi: 10.1159/000328845. [PubMed] [Cross Ref]
25. Kalfoglou AL, Boenning DA, Woolley M. Public Confidence and Involvement in Clinical Research: Symposium Summary, Clinical Roundtable, September 2000. Washington, DC: The National Academies Press; 2001.
26. Sullivan-Bolyai S, Bova C, Deatrick JA, Knafl K, Grey M, Leung K, Trudeau A. Barriers and strategies for recruiting study participants in clinical settings. West J Nurs Res. 2007 Jun;29(4):486–500. doi: 10.1177/0193945907299658. [PubMed] [Cross Ref]
27. Gamble VN. Under the shadow of Tuskegee: African Americans and health care. Am J Public Health. 1997 Nov;87(11):1773–8. [PMC free article] [PubMed]
28. Office of the Federal Register National Archives and Records Administration Code of Federal Regulations. Washington, DC: US Government Printing Office; 2002. Oct 01, [2012-12-10]. webcite 45 parts 1 to 199: public welfare
29. Tercyak KP, Johnson SB, Kirkpatrick KA, Silverstein JH. Offering a randomized trial of intensive therapy for IDDM to adolescents. Reasons for refusal, patient characteristics, and recruiter effects. Diabetes Care. 1998 Feb;21(2):213–5. [PubMed]
30. Herlihy AS, Halliday JL, Cock ML, McLachlan RI. The prevalence and diagnosis rates of Klinefelter syndrome: an Australian comparison. Med J Aust. 2011 Jan 3;194(1):24–8. [PubMed]
31. Bojesen A, Juul S, Gravholt CH. Prenatal and postnatal prevalence of Klinefelter syndrome: a national registry study. J Clin Endocrinol Metab. 2003 Feb;88(2):622–6. [PubMed]
32. Abramsky L, Chapple J. 47,XXY (Klinefelter syndrome) and 47,XYY: estimated rates of and indication for postnatal diagnosis with implications for prenatal counselling. Prenat Diagn. 1997 Apr;17(4):363–8. [PubMed]
33. Lanfranco F, Kamischke A, Zitzmann M, Nieschlag E. Klinefelter’s syndrome. Lancet. 2004;364(9430):273–83. doi: 10.1016/S0140-6736(04)16678-6. [PubMed] [Cross Ref]
34. Bojesen A, Gravholt CH. Klinefelter syndrome in clinical practice. Nat Clin Pract Urol. 2007 Apr;4(4):192–204. doi: 10.1038/ncpuro0775. [PubMed] [Cross Ref]
35. Ross JL, Roeltgen DP, Stefanatos G, Benecke R, Zeger MP, Kushner H, Ramos P, Elder FF, Zinn AR. Cognitive and motor development during childhood in boys with Klinefelter syndrome. Am J Med Genet A. 2008 Mar 15;146A(6):708–19. doi: 10.1002/ajmg.a.32232. [PubMed] [Cross Ref]
36. Rovet J, Netley C, Keenan M, Bailey J, Stewart D. The psychoeducational profile of boys with Klinefelter syndrome. J Learn Disabil. 1996 Mar;29(2):180–96. [PubMed]
37. Bojesen A, Juul S, Birkebaek NH, Gravholt CH. Morbidity in Klinefelter syndrome: a Danish register study based on hospital discharge diagnoses. J Clin Endocrinol Metab. 2006 Apr;91(4):1254–60. doi: 10.1210/jc.2005-0697. [PubMed] [Cross Ref]
38. Baker D. Chromosome errors and antisocial behavior. CRC Crit Rev Clin Lab Sci. 1972 Jan;3(1):41–101. [PubMed]
39. Schröder J, de la Chapelle A, Hakola P, Virkkunen M. The frequency of XYY and XXY men among criminal offenders. Acta Psychiatr Scand. 1981 Mar;63(3):272–6. [PubMed]
40. Witkin HA, Mednick SA, Schulsinger F, Bakkestrom E, Christiansen KO, Goodenough DR, Hirschhorn K, Lundsteen C, Owen DR, Philip J, Rubin DB, Stocking M. Criminality in XYY and XXY men. Science. 1976 Aug 13;193(4253):547–55. [PubMed]
41. Cover V. Living with Klinefelter Syndrome (47,XXY) Trisomy X (47,XXX) and 47,XYY: A guide for families and individuals affected by X and Y chromosome variations. Canada: Virginia Isaacs Cover; 2012.
42. Ross JL, Samango-Sprouse C, Lahlou N, Kowal K, Elder FF, Zinn A. Early androgen deficiency in infants and young boys with 47,XXY Klinefelter syndrome. Horm Res. 2005;64(1):39–45. doi: 10.1159/000087313. [PubMed] [Cross Ref]
43. Wikström AM, Dunkel L, Wickman S, Norjavaara E, Ankarberg-Lindgren C, Raivio T. Are adolescent boys with Klinefelter syndrome androgen deficient? A longitudinal study of Finnish 47,XXY boys. Pediatr Res. 2006 Jun;59(6):854–9. doi: 10.1203/01.pdr.0000219386.31398.c3. [PubMed] [Cross Ref]
44. Wikström AM, Painter JN, Raivio T, Aittomäki K, Dunkel L. Genetic features of the X chromosome affect pubertal development and testicular degeneration in adolescent boys with Klinefelter syndrome. Clin Endocrinol (Oxf) 2006 Jul;65(1):92–7. doi: 10.1111/j.1365-2265.2006.02554.x. [PubMed] [Cross Ref]
45. Ross JL, Zeger MP, Kushner H, Zinn AR, Roeltgen DP. An extra X or Y chromosome: contrasting the cognitive and motor phenotypes in childhood in boys with 47,XYY syndrome or 47,XXY Klinefelter syndrome. Dev Disabil Res Rev. 2009 Jul;15(4):309–17. doi: 10.1002/ddrr.85. [PMC free article] [PubMed] [Cross Ref]
46. Close S. An Exploratory Study of Physical Phenotype, Biomarkers and Psychosocial Health Parameters in Boys with Klinefelter Syndrome [dissertation] New York: Columbia University; 2011.
47. Hahn S, Janssen OE, Tan S, Pleger K, Mann K, Schedlowski M, Kimmig R, Benson S, Balamitsa E, Elsenbruch S. Clinical and psychological correlates of quality-of-life in polycystic ovary syndrome. Eur J Endocrinol. 2005 Dec;153(6):853–60. doi: 10.1530/eje.1.02024. [PubMed] [Cross Ref]
48. McCook JG, Reame NE, Thatcher SS. Health-related quality of life issues in women with polycystic ovary syndrome. J Obstet Gynecol Neonatal Nurs. 2005;34(1):12–20. doi: 10.1177/0884217504272945. [PubMed] [Cross Ref]
49. Close S. Columbia University Medical Center. 2009. [2012-12-10]. webcite 47, XXY Klinefelter Syndrome Study for Boys
50. [2012-12-10]. webcite Knowledge, support & action (KS&A)
51. RecruitSource. [2012-12-10]. webcite
52. PrivateAccess. [2012-12-10]. webcite
53. Browne K. Snowball sampling: using social networks to research non‐heterosexual women. International Journal of Social Research Methodology. 2005 Feb 2005;8(1):47–60. doi: 10.1080/1364557032000081663. [Cross Ref]
54. Eysenbach G, Wyatt J. Using the Internet for surveys and health research. J Med Internet Res. 2002;4(2):E13. doi: 10.2196/jmir.4.2.e13. [PMC free article] [PubMed] [Cross Ref]

Articles from Journal of Medical Internet Research are provided here courtesy of Gunther Eysenbach

SuperSweet—a resource on natural and artificial sweetening agents

Download PDF
This article has been cited by other articles in PMC.


A vast number of sweet tasting molecules are known, encompassing small compounds, carbohydrates, d-amino acids and large proteins. Carbohydrates play a particularly big role in human diet. The replacement of sugars in food with artificial sweeteners is common and is a general approach to prevent cavities, obesity and associated diseases such as diabetes and hyperlipidemia. Knowledge about the molecular basis of taste may reveal new strategies to overcome diet-induced diseases. In this context, the design of safe, low-calorie sweeteners is particularly important. Here, we provide a comprehensive collection of carbohydrates, artificial sweeteners and other sweet tasting agents like proteins and peptides. Additionally, structural information and properties such as number of calories, therapeutic annotations and a sweetness-index are stored in SuperSweet. Currently, the database consists of more than 8000 sweet molecules. Moreover, the database provides a modeled 3D structure of the sweet taste receptor and binding poses of the small sweet molecules. These binding poses provide hints for the design of new sweeteners. A user-friendly graphical interface allows similarity searching, visualization of docked sweeteners into the receptor etc. A sweetener classification tree and browsing features allow quick requests to be made to the database. The database is freely available at:


There are three major compounds of life: proteins; lipids and carbohydrates. The perception of sweet taste, mainly associated with advantageous food, has had an important evolutionary influence on different physiological regulation mechanisms. During human development, sugar was always luxury. In 1885 Constantin Fahlberg produced the first artificial sweetener, saccharin, and the scientific establishment was surprised by its extreme sweetness (1). Significant to this discovery was the fact that sweet taste became affordable to poor people. Following the commercial success of artificial sweeteners, a battle between the sugar and sweetener industries began (2). Saccharin was claimed to be carcinogenic in rats (3). However, it was later shown that saccharin is neither toxic nor carcinogenic in normal amounts (4), yet its reputation remains tarnished. Today, the replacement of sugar and other carbohydrates with artificial sweeteners in food is common (5) and is a general approach to prevent cavities (6,7), obesity and associated diseases such as diabetes and hyperlipidemia (8,9).

Currently, the sweet taste receptor, which is a heterodimer of two transmembrane proteins (T1R2 and T1R3) and has several different binding sites, has not been crystallized and is therefore unavailable in the Protein Data Bank (PDB) (10). Such a structure is crucial to elucidating how both small sweeteners and molecules as large as proteins bind and activate the sweet taste receptor (11). In the meantime, modeling studies can provide vital clues to these mechanisms (12). The understanding of compounds binding to the receptor is of relevance not only for the development of new artificial sweeteners but also for improving our understanding of known sweet molecules and what makes them ‘sweet’.

The first publicly available carbohydrate database was CarbBank (13), where users are able to search for carbohydrate structures, sub-structures and non-carbohydrate substituents. Wilhelm von der Lieth established the SweetDB (14), a web-based interface for glycoscientists, which was the basis for further carbohydrate tools collected in the Glycosciences portal (15) and the Glycome-DB (16) that comprises 35 000 carbohydrate sequences with a variety of query options.

There are also a number of databases that deal with glycans. GlycoBase (17) and GlycoEctractor (18) are databases that assist with interpreting high-performance liquid chromatography-glycan profiles. Tyrian Diagnostics used text-mining to develop the GlycoSuiteDB (19), which stores over 7650 glycan structures extracted from 740 papers. The Glycoconjugate Data Bank (20) provides a special tool for N-glycan primary structure verification. The connection to metabolic pathways is provided by KEGG-Glycan (21).

Although, there are a number of resources available with relation to carbohydrates, they are lacking with respect to sweetness and sweeteners. SuperSweet aims to integrate knowledge about the structure of sugars and sweetening agents with receptor binding poses, chemical properties and additional information like sweetness, approval, origin, therapeutic effect and metabolism.


The SuperSweet database was developed for researchers and dieticians and offers a user-friendly interface with helpful examples and FAQs. Currently, SuperSweet comprises more than 8000 carbohydrates, proteins, d-amino acids and artificial (synthesized) sweeteners, which were retrieved from the literature and different pre-existing data sources like Pubchem (22) and the PDB. Similarity searches extended and completed the sweetening agent data set. Besides information about the physicochemical properties of the sweet compounds, the database also offers information about the number of calories, the 3D structure, therapeutic annotations and, if detectable, the sweetness of the molecule. Structural information is available and displayed for each sweet molecule and sweet protein in the database. Moreover, the domain containing the small molecule active site of the sweet receptor was homology modeled and provided in SuperSweet (Figure 1). The small molecules were docked into the modeled binding site and the poses are also stored in the database.

Figure 1.

Homology model of the sweet taste receptor with the sweetening agent Stevioside docked. The T1R2 protomer is displayed in cartoon format and the T1R3 protomer is displayed in wireframe format with a solvent accessible surface rendered (1.2Å probe

There are different options for browsing through the database and for retrieving the data. First, the data can be retrieved by name, physicochemical properties or properties such as calories and sweetness. Secondly, the user is allowed to upload or draw a molecule using the Marvin Sketch plugin ( The query structure is compared with the entries of SuperSweet and the results presented in a table comprising molecules and a Tanimoto coefficient expressing their similarity. Thirdly, a sweet-tree is available in SuperSweet that allows the fast and easy selection of a group of sweet tasting molecules (Table 1). Here, the user can find for example, all sweet tasting proteins, or peptides or small molecules like Flavonoids. Finally, SuperSweet offers a browse section, which provides an easy way to access the SuperSweet entries by choosing different categories of molecules based on properties.

Table 1.

Organization of the Sweet-tree


Data acquisition

The sweet tasting molecules were extracted from the literature and publicly available databases like Pubchem, the PDB and MonoSaccharideDB and were filtered using different terms like ‘sweetening agents’. In the next step the data set was extended by using similarity search methods.

Homology modeling of the sweet taste receptor

A model of the large extracellular domain of the sweet taste receptor, which contains two binding sites for small molecular-weight sweeteners, was built using homology modeling. The sweet taste receptor is a class C (or metabotropic) G-protein coupled receptor and exists as a heterodimer (consisting of T1R2 and T1R3) (Figure 1). PSI-BLAST searches (23) of the PDB revealed that T1R2 and T1R3 share ∼25% sequence identity with the metabotropic glutamate receptors and are in close proximity to one another on the phylogentic tree of class C GPCRs (24). As we wanted to use the homology model of the sweet taste receptor for docking studies, it was important that we chose a template structure that is in an active (open-closed) conformation, preferably with a natural ligand bound. Accordingly, an active form of metabotropic glutamate receptor 1 (mGluR1) was used as the template (PDB code: 1EWK) (25), which also had the highest sequence coverage and the highest resolution (2.2Å) compared to other crystal structures of activated glutamate receptors. A multiple sequence alignment was created using MUSCLE (26). The alignment can be downloaded from the SuperSweet website. Homology modeling was carried out using Modeller (27) in Accelrys Discovery Studio 2.5. T1R2 was built using chain A of 1EWK (closed form) and T1R3 using chain B (open form) (12). The large insertions in T1R2 and T1R3 compared to the template were removed from the final model. Lastly, side-chain clashes were removed and the structure was minimized by carrying out 100 steps of both steepest descent and conjugant gradient minimization.

Generation of the binding poses of the small molecules

Docking of the small compounds into the homology modeled receptor was done using the docking program GOLD 4.1.1 (28). In order to define the binding site of the sweet taste receptor, the template structure (mGluR1 containing a glutamate bound to each chain) was superimposed onto the homology model of the sweet taste receptor and the glutamate molecules copied over to the homology model. The binding sites of the sweet taste receptor were then defined by using the glutamate molecules as reference ligands; all atoms within 5Å of the glutamate molecule formed the binding sites for the docking experiments. For each small molecule, 100 docking runs were performed. A previous docking study showed that the sweet taste receptor’s active site in the closed protomer is too small to host some of the larger synthetic sweeteners and is only able to host four compounds out of those tested: saccharin, alitame, aspartame and 6-Cl-tryptophan (12). Experimental work has shown that aspartame and neotame bind to the T1R2 subunit (29). In accordance with these findings, we therefore docked molecules with a molecular weight >400 kDa into T1R3 (open form) and all other molecules into the pockets of both T1R2 and T1R3. The resulting docking poses were then ranked using the GoldScore fitness function. The best scoring docking pose for each molecule can be viewed using a Jmol applet and the respective structure files are also available for download.

Conformer generation

For the small sweetening molecules, conformers were generated using the Accelrys tool. For each small molecule 20 conformers are stored and available for download on the website (30).

Similarity search

For the similarity search in the SuperSweet database, we implemented a bit vector `structural fingerprint’, which encodes the chemical and topological characteristics of a molecule. The fingerprint was pre-calculated for the small molecules of SuperSweet and is also calculated for the query structure, in order to compare it to the database entries. Open Babel implements four different fingerprints (FP2, FP3, FP4 and MACCS). Fingerprint 2 (FP2) is widely used for the comparison of small molecules and is path-based and indexes linear molecules up to seven atoms. However, this fingerprint is not ideal for use in SuperSweet due to its inability to distinguish between different ring structures and therefore between carbohydrates. To overcome this problem, we implemented a combinatorial fingerprint of fingerprint 2 and fingerprint 4. Fingerprint 4 is based on a set of SMARTS patterns and also considers functional groups. For similarity searching the Tanimoto coefficient is used, which gives values in the range of zero (fingerprints have no bits in common) to identical (all bits the same).


SuperSweet is designed as a relational database on a MySQL server. Additionally, the MyChem package ( is installed to provide a complete set of functions for handling chemical data within MySQL. Most of the functions used by MyChem depend upon Open Babel ( The structural fingerprint is implemented in Open Babel. To allow the upload or drawing of a query structure, the Marvin Sketch plugin ( was installed. For the visualization of the 3D structures Jmol ( was installed. The website is built with PHP and web access is enabled via Apache HTTP Server 2.2.


Searching for a natural sweetening agent (search field ‘Origin’) with molecular weight between 800 and 900 and sweetness above 200 returns Stevioside. Steviol glycosides like Rebaudioside A or Rubusoside are non-calorific sweeteners that are found, for instance, in sweet Chinese tea (Rubus suavissimus) and Stevia rebaudiana (31). These compounds are of research interest because advantageous effects were observed regarding cancer and blood pressure (32). These effects seem to be the result of binding to other membrane proteins (33). Clicking on the protein icon in the results table shows the docking pose of Stevioside to the sweet receptor (see Figure 1). The structure of Stevioside, the computed conformers, the best docking pose and the modeled receptor structure are downloadable from the website.

More information on Steviol glycosides can be found using the ‘Sweet-tree’ or by performing a search using the field ‘Compound name’ on the ‘Property search’ page. Clicking on the similarity search icon delivers the top 10 similar compounds.


SuperSweet compiles information on natural and artificial sweetening agents including their properties such as 3D structure, origin, sweetness, approval, calories etc. and provides hypotheses on their binding to the receptor.

Homology modeling provides a useful means of generating 3D conformations of proteins where experimental structures are not available. For this work, we generated models of the sweet taste receptor using mGluR1. The sequence identity between the receptors is rather low and is within the twilight zone of protein sequence alignments, which makes homology modeling more difficult (34). The quality of our homology model may also be affected by the fact that mGluR1 exists as a homodimer, whereas the sweet taste receptor is a heterodimer. These facts have implications for our docking experiment results due to the strong dependence of docking results on the accuracy of protein structure, especially in the binding site (35). Although GOLD has consistently been shown to be among the best performing docking algorithms in terms of the accuracy of docking poses, it is less able to distinguish the most native-like pose from all of the generated docking poses (36–38). In SuperSweet we have only made the highest scoring pose available for each docked compound and therefore the accuracy of these docking solutions should be considered in light of the aforementioned limitations in homology models and in silico docking.

Unlike the metabotropic glutamate receptors, the sweet taste receptor is predicted to have multiple binding sites (29): (i) two cavities which correspond to the Glu hosting cavities of mGluR1; (ii) a secondary binding site on the surface of the receptor for sweet proteins that corresponds to the wedge model (39) and (iii) overlapping binding sites in the seven transmembrane helix domain for the agonist cyclamate (40) and the inverse agonist lactisole (41). For this work, we only performed docking experiments to the binding sites corresponding to the Glu hosting cavities of mGluR1. These two sites differ in size due to the receptor model being in an open-closed conformation; the binding site in T1R3 is much larger as it is in an open conformation, whereas T1R2 is in a closed conformation. Therefore, we only docked compounds with a molecular weight >400 kDa into T1R3 (open form) and all other compounds into the pockets of both T1R2 and T1R3. In addition, the existence of other binding sites or alternative binding mechanisms cannot be excluded (12). Compared to mGluR1, the additional diversity of compounds binding to the sweet taste receptor and the existence of additional binding sites therefore adds further complexity to in silico docking experiments to the sweet taste receptor.

One of the future goals of SuperSweet is the integration of sugars and sweetening agents into biochemical pathway maps (including PubMed references) to better understand their different ways of metabolism and their impact on metabolic diseases and to foresee possible risk factors. After improving the similarity search and inclusion of pharmacorphore searching to find new putative sweetening agents, a sweetness prediction tool is planned to be implemented. We plan to perform additional text-mining in order to obtain information on the number of calories, sweetness and therapeutic effects of sweet compounds where missing. Docking poses of the sweet proteins to the sweet taste receptor are also planned to be integrated. Another interesting aspect would be the comparison of sweet taste perception with characteristics of sour and bitter taste perception, which is a problem in the development of artificial sweeteners.


The SuperSweet database is freely available under the url: and will be updated regularly.


Deutsche Forschungsgemeinschaft (SFB 449); Investitionsbank Berlin (IBB); Deutsche Krebshilfe; Bundesministerium für Bildung und Forschung (BMBF); European Union (EU). Funding for open access charge: DFG, BMBF and EU.

Conflict of interest statement. None declared.


The authors thank D. Kuzman and A. Chefai for manual curation of the sweetener properties and A.L. Wölke for help with the ‘Sweet tree’.


1. Merki C. Zucker gegen saccharin – zur geschichte der künstlichen süssstoffe. 1993. Campus, Frankfurt/Main/New York.
2. de la Pena C. Artificial sweetener as a historical window to culturally situated health. Ann. NY Acad. Sci. 2010;1190:159–165. [PubMed]
3. Hicks RM, Wakefield JS, Chowaniec J. Letter: Co-carcinogenic action of saccharin in the chemical induction of bladder cancer. Nature. 1973;243:347–349. [PubMed]
4. Weihrauch MR, Diehl V. Artificial sweeteners—do they bear a carcinogenic risk? Ann. Oncol. 2004;15:1460–1465. [PubMed]
5. Artificial sweeteners: no calories  sweet! FDA Consum. 2006;40:27–28. [PubMed]
6. Edgar WM. Sugar substitutes, chewing gum and dental caries—a review. Br. Dent. J. 1998;184:29–32. [PubMed]
7. Hayes C. The effect of non-cariogenic sweeteners on the prevention of dental caries: a review of the evidence. J. Dent. Educ. 2001;65:1106–1109. [PubMed]
8. Benton D. Can artificial sweeteners help control body weight and prevent obesity? Nutr. Res. Rev. 2005;18:63–76. [PubMed]
9. Drewnowski A. Intense sweeteners and energy density of foods: implications for weight control. Eur. J. Clin. Nutr. 1999;53:757–763. [PubMed]
10. Berman HM, Westbrook J, Feng Z, Gilliland G, Bhat TN, Weissig H, Shindyalov IN, Bourne PE. The Protein Data Bank. Nucleic Acids Res. 2000;28:235–242. [PMC free article] [PubMed]
11. Temussi P. The sweet taste receptor: a single receptor with multiple sites and modes of interaction. Adv. Food Nutr. Res. 2007;53:199–239. [PubMed]
12. Morini G, Bassoli A, Temussi PA. From small sweeteners to sweet proteins: anatomy of the binding sites of the human T1R2_T1R3 receptor. J. Med. Chem. 2005;48:5520–5529. [PubMed]
13. Doubet S, Albersheim P. CarbBank. Glycobiology. 1992;2:505. [PubMed]
14. Loss A, Bunsmann P, Bohne A, Loss A, Schwarzer E, Lang E, von der Lieth CW. SWEET-DB: an attempt to create annotated data collections for carbohydrates. Nucleic Acids Res. 2002;30:405–408. [PMC free article] [PubMed]
15. Lutteke T, Bohne-Lang A, Loss A, Goetz T, Frank M, von der Lieth CW. an Internet portal to support glycomics and glycobiology research. Glycobiology. 2006;16:71R–81R. [PubMed]
16. Ranzinger R, Frank M, von der Lieth CW, Herget S. a portal for querying across the digital world of carbohydrate sequences. Glycobiology. 2009;19:1563–1567. [PubMed]
17. Campbell MP, Royle L, Radcliffe CM, Dwek RA, Rudd PM. GlycoBase and autoGU: tools for HPLC-based glycan analysis. Bioinformatics. 2008;24:1214–1216. [PubMed]
18. Artemenko NV, Campbell MP, Rudd PM. GlycoExtractor: a web-based interface for high throughput processing of HPLC-glycan data. J. Proteome Res. 2010;9:2037–2041. [PubMed]
19. Cooper CA, Joshi HJ, Harrison MJ, Wilkins MR, Packer NH. GlycoSuiteDB: a curated relational database of glycoprotein glycan structures and their biological sources 2003 update. Nucleic Acids Res. 2003;31:511–513. [PMC free article] [PubMed]
20. Nakahara T, Hashimoto R, Nakagawa H, Monde K, Miura N, Nishimura S. Glycoconjugate Data Bank: Structures—an annotated glycan structure database and N-glycan primary structure verification service. Nucleic Acids Res. 2008;36:D368–D371. [PMC free article] [PubMed]
21. Hashimoto K, Kawano S, Goto S, Aoki-Kinoshita KF, Kawashima M, Kanehisa M. A global representation of the carbohydrate structures: a tool for the analysis of glycan. Genome Inform. 2005;16:214–222. [PubMed]
22. Wang Y, Xiao J, Suzek TO, Zhang J, Wang J, Bryant SH. PubChem: a public information system for analyzing bioactivities of small molecules. Nucleic Acids Res. 2009;37:W623–W633. [PMC free article] [PubMed]
23. Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z, Miller W, Lipman DJ. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 1997;25:3389–3402. [PMC free article] [PubMed]
24. Fredriksson R, Schioth HB. The repertoire of G-protein-coupled receptors in fully sequenced genomes. Mol. Pharmacol. 2005;67:1414–1425. [PubMed]
25. Kunishima N, Shimada Y, Tsuji Y, Sato T, Yamamoto M, Kumasaka T, Nakanishi S, Jingami H, Morikawa K. Structural basis of glutamate recognition by a dimeric metabotropic glutamate receptor. Nature. 2000;407:971–977. [PubMed]
26. Edgar RC. MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Res. 2004;32:1792–1797. [PMC free article] [PubMed]
27. Sali A, Blundell TL. Comparative protein modelling by satisfaction of spatial restraints. J. Mol. Biol. 1993;234:779–815. [PubMed]
28. Verdonk ML, Cole JC, Hartshorn MJ, Murray CW, Taylor RD. Improved protein-ligand docking using GOLD. Proteins. 2003;52:609–623. [PubMed]
29. Xu H, Staszewski L, Tang H, Adler E, Zoller M, Li X. Different functional roles of T1R subunits in the heteromeric taste receptors. Proc. Natl Acad. Sci. USA. 2004;101:14258–14263. [PMC free article] [PubMed]
30. Gunther S, Senger C, Michalsky E, Goede A, Preissner R. Representation of target-bound drugs by computed conformers: implications for conformational libraries. BMC Bioinformatics. 2006;7:293. [PMC free article] [PubMed]
31. Geuns JM. Stevioside. Phytochemistry. 2003;64:913–921. [PubMed]
32. Takasaki M, Konoshima T, Kozuka M, Tokuda H, Takayasu J, Nishino H, Miyakoshi M, Mizutani K, Lee KH. Cancer preventive agents. Part 8: chemopreventive effects of stevioside and related compounds. Bioorg. Med. Chem. 2009;17:600–605. [PubMed]
33. Chatsudthipong V, Muanprasat C. Stevioside and related compounds: therapeutic benefits beyond sweetness. Pharmacol. Ther. 2009;121:41–54. [PubMed]
34. Rost B. Twilight zone of protein sequence alignments. Protein Eng. 1999;12:85–94. [PubMed]
35. Bordogna A, Pandini A, Bonati L. Predicting the accuracy of protein-ligand docking on homology models. J. Comput. Chem. [Epub ahead of print, 6 July 2010]; PMID: 20607693. [PMC free article] [PubMed]
36. Kellenberger E, Rodrigo J, Muller P, Rognan D. Comparative evaluation of eight docking tools for docking and virtual screening accuracy. Proteins. 2004;57:225–242. [PubMed]
37. Kontoyianni M, McClellan LM, Sokol GS. Evaluation of docking performance: comparative data on docking algorithms. J. Med. Chem. 2004;47:558–565. [PubMed]
38. Plewczynski D, Lazniewski M, Augustyniak R, Ginalski K. Can we trust docking results? Evaluation of seven commonly used programs on PDBbind database. J. Comput. Chem. [Epub ahead of print, 1 September 2010]; PMID: 20812323. [PubMed]
39. Temussi PA. Why are sweet proteins sweet? Interaction of brazzein, monellin and thaumatin with the T1R2-T1R3 receptor. FEBS Lett. 2002;526:1–4. [PubMed]
40. Jiang P, Cui M, Zhao B, Snyder LA, Benard LM, Osman R, Max M, Margolskee RF. Identification of the cyclamate interaction site within the transmembrane domain of the human sweet taste receptor subunit T1R3. J. Biol. Chem. 2005;280:34296–34305. [PubMed]
41. Jiang P, Cui M, Zhao B, Liu Z, Snyder LA, Benard LM, Osman R, Margolskee RF, Max M. Lactisole interacts with the transmembrane domains of human T1R3 to inhibit sweet taste. J. Biol. Chem. 2005;280:15238–15246. [PubMed]

Articles from Nucleic Acids Research are provided here courtesy of Oxford University Press

Assessment of dose homogeneity in conformal interstitial breast brachytherapy with special respect to ICRU recommendations

Download PDF



To present the results of dose homogeneity analysis for breast cancer patients treated with image-based conformal interstitial brachytherapy, and to investigate the usefulness of the ICRU recommendations.

Material and methods

Treatment plans of forty-nine patients who underwent partial breast irradiation with interstitial brachytherapy were analyzed. Quantitative parameters were used to characterize dose homogeneity. Dose non-uniformity ratio (DNR), dose homogeneity index (DHI), uniformity index (UI) and quality index (QI) were calculated. Furthermore, parameters recommended by the ICRU 58 such as minimum target dose (MTD), mean central dose (MCD), high dose volume, low dose volume and the spread between local minimum doses were determined. Correlations between the calculated homogeneity parameters and usefulness of the ICRU parameters in image-based brachytherapy were investigated.


Catheters with mean number of 15 (range: 6-25) were implanted in median 4 (range: 3-6) planes. The volume of the PTV ranged from 15.5 cm3 to 176 cm3. The mean DNR was 0.32, the DHI 0.66, the UI 1.49 and the QI 1.94. Related to the prescribed dose, the MTD was 69% and the MCD 135%. The mean high dose volume was 8.1 cm3 (10%), while the low dose volume was 63.8 cm3 (96%). The spread between minimum doses in central plane ranged from −14% to +20%. Good correlation was found between the DNR and the DHI (R2=0.7874), and the DNR correlated well with the UI (R2=0.7615) also. No correlation was found between the ICRU parameters and any other volumetric parameters.


To characterize the dose uniformity in high-dose rate breast implants, DVH-related homogeneity parameters representing the full 3D dose distributions are mandatory to be used. In many respects the current recommendations of the ICRU Report 58 are already outdated, and it is well-timed to set up new recommendations, which are more feasible for image-guided conformal interstitial brachytherapy.

Keywords: breast cancer, homogeneity, dose-volume histogram, image-based brachytherapy


In interstitial brachytherapy (BT), the non-homogeneous dose distribution around the radioactive sources is mainly determined by inverse square law. Other factors add only little modifications to this geometrical phenomenon. In the immediate proximity of the sources there are always regions of high dose, but with appropriate source distribution regions with low dose gradient can be attained, and in proper implants, the high dose volumes are relatively small. Historically, different parameters have been defined to characterize the dose homogeneity in BT [14]. The International Commission on Radiation Units and Measurements (ICRU) published the ICRU Report 58 in 1997 which deals with specification of dose homogeneity in interstitial BT [5]. For a reporting purpose it is recommended to use homogeneity parameters which have been validated in classical low dose rate (LDR) BT. However, in modern image-guided, dose optimized high-dose-rate (HDR) BT in which stepping-source remote afterloading equipment is used for irradiation, the practical applicability of these parameters is questionable. Boost dose after whole breast irradiation as well as accelerated partial breast irradiation (APBI) can be delivered with image-guided BT where conformal dose distribution can be achieved with optimized dose distribution [69]. Among the APBI techniques, the longest experience is present in multi-catheter based interstitial BT [79]. Now, follow-up data of up to 12 years are already available for HDR interstitial breast BT with comparable results to the WBI in terms of safety and efficacy [8]. In 2004, a European multicentre Phase III clinical trial was initiated by the Breast Cancer Working Group of the GEC-ESTRO to investigate the efficacy of the APBI [9]. Our institution actively participated in this study.

The purpose of this paper is to present the results of a detailed analysis on dose homogeneity of dose distributions in treatment plans made for our patients enrolled into the GEC-ESTRO trial and treated with interstitial BT. Furthermore, to investigate the suitability of the ICRU recommendations for dose uniformity in image-based conformal interstitial BT.

Material and methods

Dose plans of forty-nine patients were evaluated with respect to dose homogeneity. All patients were treated with microSelectron V2 HDR afterloader (Nucletron BV, Weenendaal, The Netherlands), and the used planning system was the Nucletron’s Plato Brachytherapy v.14.6. The details of our planning and implant techniques have been published elsewhere [10]. We used pre- and post-implant CT imaging for catheter placements and treatment planning. Following geometrical and graphical optimization, the dose was normalized to basal dose points and an isodose line was individually selected for dose prescription in order to obtain at least 90% of target volume coverage. The prescribed dose (PD) was 30.1 Gy delivered by 7×4.3 Gy, 2 fractions daily. Quantitative evaluation of dose plans was performed with dose volume histograms (DVHs). To characterize the homogeneity of dose distributions, the most common DVH based quality indices and parameters recommended by the ICRU were calculated. Descriptive statistics was calculated and correlation analysis between the indices and parameters was performed. Volumetric homogeneity parameters used for calculations were as follows:

Dose non-uniformity ratio (DNR)

The DNR is a simple and easy to interpret parameter for quantitative analysis of dose homogeneity in interstitial implants. The DNR is the ratio of the high dose volume to the reference dose volume [3]. The reference dose volume is the volume that receives dose equal or greater than PD, and the high dose volume is the volume that receives 1.5 times PD or more. The optimal dose distribution in terms of dose uniformity can be achieved at the minimum DNR value.

Dose homogeneity index (DHI)

The concept of DHI is similar to DNR, though different definitions exist in the literature [1113]. Sometimes it is used as a complementary parameter to the DNR (DHI = 1 – DNR). It can be calculated only for the implant geometry and can also be related to the volume of the PTV. In the latter case it is called relative homogeneity index (HI). In this paper we used the definition as follows: DHI=(V100 – V150)/V100. Where, V100 and V150 is the relative volume of the PTV in percent irradiated at least by the 100% and 150% of the PD, respectively.

Uniformity index (UI)

The UI is calculated from the “natural” volume dose histogram (NVDH) [1]. In the NVDH the “u” parameter is defined as –3/2 power of dose (D-3/2), and the volume (V) per unit “u” (dV/du) is plotted versus the “u” parameter. With this transformation the inverse square law is suppressed, and from this follows that for a point source the NVDH is a horizontal line. For a real implant, there is a peak on the graph which is graphical representation of the dose uniformity. The narrower the peak, the more uniform the dose distribution is. Evaluation of other peak parameters such as width, position and contained volume, in relation to treatment dose permits to define other quantitative volume-dose parameters such as UI and QI.

Per definition,

equation M1


where TD is the treatment dose (or PD) and HD (high dose) is dose value at dV/du that is half way between the (dV/du)max (peak dose, PkD) and the asymptotic value of dV/du as u → infinity (see HD definition in Fig. 1). The UI depends on the prescribed dose, thus it can be used to compare implants with the same dose prescription only.

Fig. 1

Natural volume-dose histogram for a breast implant. The arrows shows how the LD and HD are defined. The LD and HD is used to define the QI and UI, respectively

Quality index (QI)

The formula of QI is similar to UI, but instead of the treatment dose (TD) the low dose (LD) is used, where LD is dose value at dV/du that is half way between the (dV/du)max (peak dose, PkD) and the asymptotic value of dV/du as u → 0 (see LD definition in Fig. 1). Since treatment dose is excluded from the formula, QI is independent on the prescribed dose.

ICRU Report 58 recommendations

The ICRU recommends using the following parameters in interstitial BT: Minimum Target Dose (MTD) – minimum dose at the periphery of the clinical target volume, which in most cases practically coincides with the PTV, Mean Central Dose (MCD) – arithmetic mean of the local minimum doses between sources in the central plane (same as basal dose in the Paris system), High dose volume – volume encompassed by the isodose corresponding to 150% of the MCD, Low dose volume – volume within the clinical target volume encompassed by the isodose corresponding to 90% of the PD (corresponds to V90).

For high dose volume and low dose volume the maximum dimension of the volumes in the calculated planes should be reported. Dose uniformity parameters are the mean spread between the local minimum doses in the central plane (maximal±percentage deviations of the individual minimum doses from the MCD) and the MTD/ MCD (ratio of MTD and MCD).


The median number of implanted catheters was 15 (range: 6-25) in a median of 4 (range: 3-6) planes. The mean volume irradiated by the PD was 78.8 cm3 (range: 23.2-209.5 cm3). The volume of the PTV ranged from 15.5 cm3 to 176 cm3 with a mean of 66.4 cm3. The volumetric dose homogeneity parameters are shown in Table 1. In 6 out of 49 cases (12%) the DNR value was higher than 0.35 which was the upper limit in the study. But, this was always accepted in order to obtain proper dose coverage. The dose homogeneity inside the PTV is characterized by 0.66 (range: 0.50-0.76) as a mean of the DHI.

Table 1

Volumetric homogeneity parameters for 49 HDR breast implants

Table 2 shows calculated parameters recommended by the ICRU. The wide range of the MTD (53-92%) indicates weakness of the use of this parameter in conformal BT. The average of the MCDs is 135%, which means that the mean isodose selected for dose prescription was 74% (range: 69-85%). The mean volume irradiated by 1.5 times MCD was 10% (range: 6-36%) which corresponds to absolute volume of 8.1 cm3 (range: 3.4-21.4 cm3). The mean low dose volume (96%) was close to 100% corresponding to 63.8 cm3. The mean deviation in local mean minimum doses from the MCD was 14% in negative and 20% in positive direction. The largest deviation was –25% and +61%. The minimum dose in the target related to the MCD (MTD/MCD) was quite low with 0.51 (range: 0.37-0.69) value.

Table 2

Homogeneity parameters recommended by the ICRU Report 58 for 49 breast implants

Figure 2 shows the correlation between DNR and DHI. Although, the former relates to the implant geometry and the latter to the PTV, the correlation is quite good (R2=0.7874). The UI also correlates with the DNR, which is presented in Fig. 3. No correlation (R2 <0.5) was found between the ICRU parameters (spread in individual minimum doses, MTD/MCD, low dose volume, high dose volume) and any other volumetric parameters (DNR, DHI, UI).

Fig. 2

Correlation between the DHI and DNR
Fig. 3

Correlation between the UI and DNR


In interstitial BT, the classical Paris system has been successfully used clinically for different treatment sites for decades [14]. One of the advantages of the Paris system is that following its rules the resulting dose distribution will be always homogeneous. Although, it was originally based on LDR wire sources, its application is also possible with a HDR stepping source, when uniform dwell times are used [15]. In a previous study, comparing different dosimetry systems we found that the most homogeneous dose distributions occurred in the Paris dosimetry system and in the geometrical optimization [16]. For both systems the mean DNR was 0.25. The clinical availability of dose optimization algorithms and recent evolution of image-based brachytherapy have highlighted the limitations of the Paris system [17]. With 3D imaging, the exact definition of the PTV is possible, and this calls for tailoring the reference isodose surface to the PTV. However, good dose coverage sometimes can be achieved only with deterioration of dose homogeneity [16, 18].

At the time of publication of the ICRU Report 58, conformal interstitial BT was not widely available. This is well reflected by the recommended parameters which can be effectively used in projection-based classical implants. Use of an implant related parameters and point doses is recommended, and DVH is mentioned only as an additional representation of dose distributions. This is understandable, since at that time individual computerized treatment planning was not common. Without 3D volume calculation, the dimensions of the high dose volume in different planes have to be determined as per the recommendations. In current planning systems, however, calculation of the high dose volume can be easily performed from the DVH. In LDR BT or in HDR stepping source BT with uniform dwell times, the volume irradiated by 1.5 times MCD (high dose volume according to the ICRU) can be approximated by the dimensions measured in three planes, since the high dose region closely follows the catheters. But, in conformal BT the source dwell times can be very different due to optimization algorithms. From this follows, that the high dose volume will be irregular (bumpy) and its size can not be estimated with dimensions measured only in three planes. This is demonstrated in Fig. 4, where 3D representation of the high dose volume is shown in uniform and various dwell times in a two-plane breast implant. In the latter case, the dwell times were determined by optimization algorithm. It is evident from the images that knowing the dimensions in 2D planes only, can not be equivalent to calculation of the full 3D volume if the source dwell times are not uniform.

Fig. 4

3D representation of high dose volume according to the ICRU Report 58 in a breast implant planned A) without optimization, B) with geometrical and graphical optimization

In BT the dose inhomogeneity is unavoidable and it is particularly important in cases of breast implants, where all the breast tissue can be considered equally at risk for developing late side effects (e.g. fibrosis or fat necrosis). Wazer et al. [12] found a significant relationship between dose homogeneity and cosmetic outcome in interstitial LDR boost breast implants. With higher value of DHI they observed less late fibrosis. In another study from the same department, no clear statistical correlation between dose homogeneity and complication risk was found at sole HDR brachytherapy treatment for early-stage breast cancer [19]. In a study with LDR implants the probability of excellent cosmetic outcome linearly increased with DHI [20]. Vicini et al. [11] reported the DHI of 0.89–0.90 calculated for the implants of five patients. In the study of Das et al. [13] the DHI ranged from 0.46 to 0.85 with a mean of 0.73. Converting these values into DNR, the range will be from 0.15 to 0.54 with a mean of 0.27. Recently, a new dose volume uniformity index has been proposed where all volume elements irradiated by higher than the prescribed dose is taken into account [21].

The parameters for dose uniformity recommended by the ICRU relates to 2D dose distributions and point doses. Our results demonstrated that this simple representation of dose homogeneity did not correlate with volumetric parameters in HDR implants. The spread in individual minimum doses in the central plane may describe dose homogeneity in that plane, but the degree of homogeneity in the whole volume can be very different. In our study, there was no correlation between the deviations in the midpoint doses between the catheters in the central plane and volumetric parameters (DNR, DHI). The explanation for this is that in optimized dose plans the dose distributions in planes parallel to central plane can be unrelated to each other, not like in classic LDR implants or HDR implants with uniform source dwell times. Therefore, the central plane is no longer representative of the implant, as it was before. Nowadays, 3D assessments of dose distribution is mandatory with volumetric parameters to characterize the dose homogeneity.


In the era of image-guided conformal interstitial BT, the recommendations of the ICRU Report 58 seem to be outdated in many respects. The progress in imaging and dose optimization algorithms has recently made conformal BT as a routine procedure in many institutions. Considering this, it is mandatory to use DVH-related homogeneity parameters representing fully the 3D dose distributions. To decide which parameters have clinical significance requires more studies with clinical validation of their correlation with treatment outcome and side effects. In order to report the treatments in consistent way, new recommendations from international bodies and/or professional societies are highly awaited.


This paper is partially supported by the Sectoral Operational Programme Human Resources Development, financed from the European Social Fund and by the Romanian Government under the contract number POSDRU/89/1.5/S/60782.


1. Anderson LL. A “natural” volume dose histogram for brachytherapy. Med Phys. 1986;13:898–903. [PubMed]
2. Wu A, Ulin K, Sternick ES. A dose homogeneity index for evaluating Ir-192 interstitial breast implants. Med Phys. 1988;15:104–107. [PubMed]
3. Saw CB, Suntharalingam N, Wu A. Concept of dose nonuniformity in interstitial brachytherapy. Int J Radiat Oncol Biol Phys. 1993;26:519–527. [PubMed]
4. Wong VYW, Leung TW, Wong CM. Relative dose uniformity assessment in interstitial implants. Int J Radiat Oncol Biol Phys. 1999;44:1179–1184. [PubMed]
5. ICRU. Bethesda, USA: ICRU; 1997. Dose and volume specification for reporting interstitial therapy, ICRU Report 58.
6. Polgár C, Jánváry L, Major T, et al. The role of high-dose-rate brachytherapy boost in breast-conserving therapy: Long-term results of the Hungarian National Institute of Oncology. Rep Pract Oncol Radiother. 2010;15:1–7.
7. Offersen BV, Overgaard M, Kroman N, et al. Accelerated partial breast irradiation as part of breast conserving therapy of early breast carcinoma: A systematic review. Radiother Oncol. 2009;90:1–13. [PubMed]
8. Polgár C, Major T, Fodor J, et al. Accelerated partial-breast irradiation using high-dose-rate interstitial brachytherapy: 12-year update of a prospective clinical study. Radiother Oncol. 2010;94:274–279. [PubMed]
9. Polgár C, Strnad V, Major T. Brachytherapy for partial breast irradiation: the European experience. Semin Radiat Oncol. 2005;15:116–122. [PubMed]
10. Major T, Fröhlich G, Lövey K, et al. Dosimetric experience with accelerated partial breast irradiation using image-guided interstitial brachytherapy. Radiother Oncol. 2009;90:48–55. [PubMed]
11. Vicini FA, Kestin LL, Edmundson GK, et al. Dose-volume analysis for quality assurance of interstitial brachytherapy for breast cancer. Int J Radiat Oncol Biol Phys. 1999;45:803–810. [PubMed]
12. Wazer DE, Kramer B, Schmid C, et al. Factors determining outcome in patients treated with interstitial implanatation as a radiation boost for breast conservation therapy. Int J Radiat Oncol Biol Phys. 1997;39:381–393. [PubMed]
13. Das RK, Patel R, Shah H, et al. 3D CT-based high-dose-rate breast brachytherapy implants: treatment planning and quality assurance. Int J Radiat Oncol Biol Phys. 2004;59:1224–1228. [PubMed]
14. Pierquin B, Dutreix A, Paine CH, et al. The Paris System in interstitial radiation therapy. Acta Radiol Oncol. 1978;17:33–48. [PubMed]
15. van der Laarse R. The stepping source dosimetry system as an extension of the Paris system. In: Mould RF, Battermann JJ, Martinez AA, Speiser BL, editors. Brachytherapy from radium to optimization, Veenendaal. The Netherlands: Nucletron BV; 1994. pp. 319–330.
16. Major T, Fodor J, Takacsi-Nagy Z, et al. C. Evaluation of HDR interstitial breast implants planned by conventional and optimized CT-based dosimetry systems with respects to dose homogeneity and conformality. Strahlenther Onkol. 2005;181:89–96. [PubMed]
17. Hennequin C, Mazeron JJ, Chotin G. How to use the Paris system in the year 2001? Radiother Oncol. 2001;58:5–6. [PubMed]
18. Cholewka A, Szlag M, Slosarek K, et al. Comparison of 2D- and 3D-guided implantation in accelerated partial breast irradiation (APBI) J Contemp Brachyther. 2009;1:207–210.
19. Wazer DE, Lowther D, Boyle T, et al. Clinically evident fat necrosis in women treated with high-dose-rate brachytherapy alone for early-stage breast cancer. Int J Radiat Oncol Biol Phys. 2001;50:107–111. [PubMed]
20. Kramer BA, Arthur DW, Ulin K, et al. LDR – probability of excellent cosmetic outcome linearly increased with DHI. Radiology. 1999;213:61–66. [PubMed]
21. Prabhakar R. Dose volume uniformity index: a simple tool for treatment plan evaluation in brachytherapy. J Contemp Brachyther. 2010;2:71–75.

Articles from Journal of Contemporary Brachytherapy are provided here courtesy of Termedia Publishing

Increased Carotid Intima-Media Thickness and Reduced Distensibility in Human Class III Obesity: Independent and Differential Influences of Adiposity and Blood Pressure on the Vasculature

Download PDF
Guillermo López-Lluch, Editor


Carotid intima-media-thickness (cIMT) and carotid distensibility (distensibility), structural and functional properties of carotid arteries respectively, are early markers, as well as strong predictors of cardiovascular disease (CVD). The characteristic of these two parameters in individuals with BMI>40.0 kg/m2 (Class III obesity), however, are largely unknown. The present study was designed to document cIMT and distensibility in this population and to relate these to other factors with established association with CVD in obesity. The study included 96 subjects (65 with BMI>40.0 kg/m2 and 31, age- and gender-matched, with BMI of 18.5 to 30.0 kg/m2). cIMT and distensibility were measured by non-invasive high resolution ultrasonography, circulatory CD133+/KDR+ angiogenic cells and endothelial microparticles (EMP) by flow cytometry, and plasma levels of adipokines, growth factors and cytokines by Luminex immunoassay kits. The study results demonstrated increased cIMT (0.62±0.11 mm vs. 0.54±0.08 mm, P = 0.0002) and reduced distensibility (22.52±10.79 10−3kpa−1 vs. 29.91±12.37 10−3kpa−1, P<0.05) in individuals with BMI>40.0 kg/m2. Both cIMT and distensibility were significantly associated with traditional CVD risk factors, adiposity/adipokines and inflammatory markers but had no association with circulating angiogenic cells. We also demonstrated, for the first time, elevated plasma EMP levels in individuals with BMI>40.0 kg/m2. In conclusion, cIMT is increased and distensibility reduced in Class III obesity with the changes predominantly related to conventional CVD risk factors present in this condition, demonstrating that both cIMT and distensibility remain as CVD markers in Class III obesity.


Carotid intima-media-thickness (cIMT) and carotid distensibility (distensibility) represent structural and functional properties of carotid arteries respectively. Both increased cIMT, a noninvasive measure of subclinical atherosclerosis, and reduced distensibility, an indicator of regional artery stiffness, are independent predictors of future cardiovascular events [1], [2]. Importantly, a combined assessment of the two allows for a better analysis of the individual atherosclerotic burden and improved prediction of aortic atherosclerosis [3].

Increased cIMT or decreased distensibility has been linked to hypertension [4], diabetes mellitus [5] and obesity [6][10], determinant risk factors for cardiovascular disease (CVD) [11][14]. The occurrence of these three co-morbidities is linked with chronic low-grade inflammation. Furthermore insulin resistance present in obesity is believed to be a principal contributor to this link. The inter-relation between adipogenesis, inflammation, insulin resistance, hypertension and diabetes mellitus remains a current focus of obesity research. Nevertheless higher CVD incidences are evident in hypertensive and/or diabetic obese compared to non-obese counterparts. The prevalence of obesity is rising at an alarming rate worldwide. Moreover the prevalence of Class III obesity, defined as BMI≥40.0 kg/m2, is increasing at an even steeper rate [15], [16]. cIMT and distensibility in Class III obesity, however, are largely undocumented with only two papers providing both cIMT and distensibility data in people with BMI≥40.0 kg/m2 [17], [18]. Similarly, little information is available in Class III obesity on novel biomarkers of CVD, such as circulatory angiogenic cells [19] or endothelial microparticles [20].

The main objective of this study was therefore to document cIMT and distensibility in Class III obese subjects compared with a non-obese cohort, and to examine and compare traditional CVD risk factors (CVRF) and novel CVD biomarkers between the two populations. We hypothesized that cIMT and distensibility remain useful as CVD markers in the severely obese population despite technical difficulties that may be encountered and verified this by determining the association of cIMT and distensibility with other established CVRF in Class III obesity.

Materials and Methods

Ethics Statement

The study protocol was approved by the institutional ethics committee of Alfred Healthcare (#158/06), and informed written consent was obtained from each participant.

Study Population and Design

A total of 96 subjects (31 non-obese controls: BMI 18.5 to 30.0 kg/m2 and 65 class III obesity: BMI>40.0 kg/m2) were included in the study. Class III obesity subjects were recruited via the Obesity Research Groups at Monash University while the age- and gender-matched non obese were from the Baker IDI BioBank database. Exclusion criteria were known coronary artery disease, cardiac failure, vascular brain disease, peripheral obstructive artery disease, significant renal or hepatic dysfunction and pregnancy. Subjects with current or past history of multiple myeloma, blood dyscrasia or any form of leukemia or lymphoma were also excluded.

All individuals underwent a physical examination and had their medical histories recorded. In brief, participants were measured for height, weight, waist and hip circumferences and blood pressure. 30 ml of peripheral blood was drawn for routine blood tests following a 12 hr fast and also analyzed for levels of plasma adipokines, growth factors and cytokines, circulating angiogenic cells (CD133+/KDR+ PBMCs & Hill-CFU) and endothelial microparticles (EMP). Routine blood tests were performed by the Alfred Pathology Department including a full blood count, hsCRP, glucose and a lipid profile (HDL, LDL, total cholesterol and triglycerides). cIMT and distensibility were examined using non-invasive high resolution ultrasonography.

The measurement of cIMT and distensibility were compared between the two groups. The associations of cIMT or distensibility with traditional CVRF (age, gender, BP, glucose and lipids etc) and adiposity/adipokines (BMI, waist:hip, adiponectin and leptin) were examined, as were their respective associations with inflammatory markers and circulating angiogenic cells. Plasma levels of EMP were measured in randomly selected subpopulations including both males and females (Class III obese = 15 and non-obese = 16) to determine vascular inflammation and integrity.

Carotid Imaging and Measurement

The left and right common carotid arteries proximal to the carotid bifurcation were imaged through non-invasive high resolution ultrasonography using a Philips iE33 ultrasound system (Philips, Bothell, WA, USA) with a 11–3 MHz linear array transducer while the subject was at rest in a supine position. Briefly, the carotid arteries were imaged in longitudinal sections, 0–2 cm proximal to the carotid artery bifurcation, focusing on the far wall of the vessel. Two 10-second loops were captured for each of the left and right arteries and stored for offline analysis. cIMT was defined as the distance between the intima-lumen interface and the media-adventitia interface, and measured at end diastole (as determined from the simultaneous electrocardiogram recordings) over a 10 mm long portion of the vessel wall between 0–1 cm proximal to the carotid bulb. Diastolic and systolic diameters, for distensibility calculation, were determined as the smallest and largest diameter values during a cardiac cycle. An average of three measurements from consecutive cardiac cycles from each of the left and the right carotid artery was made, and the average of the left and right arteries was used for the final analysis. All measurements were conducted independently using an automatic edge detection system (Philips QLAB version 7.0) by two observers blinded to all participant information.

Carotid distensibility was calculated as (2Δd/ds)/ΔP in 10−3•kPa−1, where Δd is carotid internal diameter change between systole and diastole, ds is carotid systolic diameter and ΔP is pulse pressure [21].

Determination of Plasma Adipokines and Cytokines

Plasma samples were collected and stored at −80°C until use. Plasma levels of adiponectin, leptin, IL-10 and SDF-1 were measured using Luminex immunoassay kits (Millipore, USA) as per manufacturer’s instruction. Briefly, the appropriate adipokines or cytokine standards, plasma samples (25 µL), and fluorescent conjugated, antibody-immobilized beads were added to wells of a pre-wet filtered plate and then incubated in dark overnight at 4°C. The following day, the plate was washed twice with wash buffer and then incubated with secondary detection antibody for 1 hr, followed by subsequent incubation with streptavidin-PE for 30 min. After the plate was washed twice again with wash buffer, it was run on the Luminex system (BioRad) with the addition of sheath fluid. Concentrations of different analytes in the plasma samples were determined by using respective standard curves generated in the assays.

Measurement of Circulating Angiogenic Cells

Enumeration of peripheral blood CD133+/KDR+ PBMCs by flow cytometry

Peripheral blood mononuclear cells (PBMCs), isolated from fresh venous blood by ficoll-gradient centrifugation, were used to quantitate the number of CD133+/KDR+ PBMCs by flow cytometry. In brief, 100 µl of PBMCs (1.5×107/ml) was incubated with Fc-γ receptor blocking agent followed by 30 min incubation on ice with antibodies against human CD133 (PE-conjugated, Miltenyi Biotec, Germany) and VEGFR-2 (KDR) (APC-conjugated, R&D Systems, USA). PE- and APC-conjugated mouse IgG from the same manufacturers served as isotype controls. Following incubation, cells were washed with PBS and then fixed in 1% paraformaldehyde. Flow cytometry acquisition was performed on BD FACSCalibur™ using appropriate settings excluding debris and platelets as shown in Figure 1-A1 & Figure 1-B1. 106 events per sample were collected within the R1 monocyte gates. Cells positive for both CD133 and KDR (upper right quadrants of the FL2-FL4 plots as shown in Figure 1-A2 & Figure 1-B2) were characterized as angiogenic cells. Results were expressed as percentage of CD133+KDR+PBMCs/PBMCs. Analysis was carried out by a blinded randomized approach in regard to patient profiles using the FlowJo software (Tree Star Inc, USA).

Figure 1

Illustration of FACS gating analysis of angiogenic cells (AC133+/KDR+PBMCs).

Hill-CFU assay

The ability to clonally expand and generate colonies in an endothelial-specific medium is considered a key functional feature of angiogenic cells. The Hill-CFU assay was performed using the commercially available kit, EndoCultTM liquid medium Kit (Stem cells Technologies, USA) and used as per the manufacturer’s instructions. In brief, 5×106 ficoll-isolated PMNCs were resuspended with 2 ml EndoCult medium and plated in a well of fibronectin-coated 6-well plates (BD Biosciences, USA), which were incubated for two days at 37°C, 5% CO2 with ≥95% humidity. After two days the non-adherent cells were harvested, counted and plated as at a density of 1×106 cells per well onto a 24-well fibronectin-coated plate, which was then incubated at 37°C for another 3 days. Hill-CFUs, characterized by a central cellular cluster surrounded by emerging spindle-shaped cells were counted at day 5 in 24-well plates in a minimum of 3 wells per subject and the average count was recorded. Results were expressed as number of colony per well.

Isolation and Identification of Circulating EMPs

EMPs are defined as CD31+/CD41 particles sized between 0.1–1 µm in platelet-depleted plasma. They were determined by the analysis for the expression of surface antigens by flow cytometry. In brief, 500 µl of completely thawed plasma was centrifuged at 16000 g for 5 min at 4°C to deplete platelets or any cell debris. The top 450 µl of plasma was transferred into a fresh tube, which was centrifuged again at 16000 g for 30 min at 4°C. The top 250 µl plasma was carefully removed and the remaining 200 µl vortexed and used for FACS analysis. Following a 15 min incubation with 50 µl Fc-γ receptor blocking agent (Miltenyi Biotec, Germany) at room temperature to reduce non-specific binding, half of the treated plasma was incubated with antibodies against human CD31 (Alexa647-conjugated, BD Biosciences, USA) and CD41 (PE-conjugated, BD Biosciences, USA). The other half was incubated with Alexa647- and PE-conjugated mouse IgG from the same manufacturer served as isotype controls. At the end of 20 min incubation, 300 µl of double filtered 1% Formaldehyde/0.2% FBS/PBS (filtered through a 0.2 and then a 0.1 µm membrane filter before use) was added for fixation and 50 µl of diluted calibration beads (BD Biosciences, USA) was added for EMP calculation and size reference. Each sample and its corresponding control were counted on BD FACSCalibur™ (BD Biosciences, USA) for 5 min.

For flow cytometry counting, EMP gate (R2 as shown in Figure 2-B) was pre-defined using commercial beads sized at 0.1 and 1 µm (Sigma-Aldrich, USA). Only events included within this gate were further analysed for fluorescence signal as shown in Figure 2-D. For EMP enumeration, a formula was used based on the concentration of the added calibration beads [22], which discriminated themselves from the EMP population on the FSC-SSC cytogram (R1 as shown in Figure 2-B and Figure 2-C). All counting data were then processed with a blinded randomized approach using BD CellQuest Pro (BD Biosciences, USA). Results are presented as number of CD31+/CD41 EMP per µl of plasma.

Figure 2

Elevated levels of circulating CD31+/CD41 EMP in obesity.

Statistical Analysis

Logarithmic transformations were applied if appropriate to skewed data following histogram analyses and Kolmogorov-Smirnov test. Transformed data are expressed as geometric mean (95% CI) and non-transformed data are expressed as mean ± SD. Comparisons between the non-obese and Class III obese groups were performed by two tailed Student’s t-test. Bivariate correlation analysis of adiposity (BMI and waist:hip), carotid variables (cIMT and distensibility) and blood pressure (SBP) was performed to define each crude association with other variables measured. Multivariable linear regression models were then constructed with the use of important covariates concluded from correlation analysis (P<0.1), in a hierarchal fashion, to elucidate independent determinants of cIMT and distensibility. Model 1 was adjusted for traditional CVRF (age, SBP, BP-med, fasting glucose and triglycerides), while Model 2 for traditional CVRF and adiposity/adipokines (BMI, adiponectin and leptin), and Model 3 for traditional CVRF, adiposity/adipokines and inflammatory markers (hsCRP, IL10 and WBC). Multivariable analysis was also repeated with no BP adjustment to further assess reliability of independent association between SBP and distensibility, since distensibility is a derivative parameter related to pulse pressure. For the same purpose, multivariable analysis of distensibility was again carried out separately in non-hypertensive and hypertensive subjects. In addition, multivariable analysis of SBP was performed as well. All statistical analyses were performed with SPSS version 12.0 for Windows and a probability value P<0.05 was considered statistically significant.


The demographic, anthropometric, clinic and laboratory characteristics of the 96 subjects included in the study are shown in Table 1. Except for age, gender and smoking history, the 31 non-obese and 65 Class III obese subjects presented as two distinctive phenotypes in respective to all parameters related to traditional CVRF, adiposity, and plasma levels of circulating adipokines as well as inflammatory markers. All traditional CVRF measured were significantly worse in the Class III obese group. The heavier atherosclerotic burden in the Class III obese subjects was demonstrated both through increased cIMT (P = 0.0002) and reduced distensibility (P<0.05) in comparison to their age- and gender-matched non-obese counterparts (Figure 3-A & Figure 3-B). Furthermore, significantly higher levels of circulating EMP (P = 0.02) indicated vascular inflammation and a compromised integrity of vascular endothelium in Class III obese subjects (Figure 2-A).

Figure 3

Increased cIMT (A) and reduced distensibility (B) in obesity.
Table 1

Anthropometric, clinical and biochemical characteristics.

As expected, significantly elevated plasma leptin and suppressed adiponectin, a typical adipokine phenotype of obesity, were shown in Class III obese subjects (Table 1), as was the inflammatory profile of much higher levels of plasma hsCRP but lower IL10. Both measures of adiposity, BMI and waist:hip, closely correlated to status of metabolic syndromes (blood pressure, glucose, type 2 diabetes, HDL and triglycerides) and inflammation (Table 2).

Table 2

Bivariate correlation between BMI, Waist:Hip, SBP, cIMT and CD with other covariates.

On the other hand, there was no difference in the number of circulatory angiogenic cells (CD133+KDR+PBMC) between obese and non-obese populations, despite significantly higher colony-forming capacities (Hill-CFU) in Class III obese subjects (Table 1). Plasma levels of SDF-1, a critical cytokine mobilizing angiogenic cells, were nearly doubled (P = 0.003) in subjects with BMI>40 kg/m2.

cIMT significantly correlated with age, BMI, waist:hip ratio, hsCRP, and status of the metabolic syndrome (Table 2). These correlations were also true for distensibility. As well, distensibility was also linked to BP-medication and plasma levels of leptin and adiponectin.

Subjects with hypertension were defined, in this study, for those who either presented with elevated blood pressure (≥140/90 mmHg) at the time of BP measurement or were taking antihypertensive medications (BP-med) or had a history of hypertension. Blood pressure was significantly elevated in Class III obese subjects, and multivariate regression analysis revealed that SBP was independently associated with BMI (β = 0.879, P<0.0001) and plasma levels of adiponectin (β = −0.304, P = 0.049) in this cohort. Multivariate regression analysis of cIMT and distensibility, therefore, was performed with and without BP adjustment (Table 3). The results showed that age and BMI were independently associated with cIMT regardless of BP adjustment (Table 3). In contrast, carotid distensibility was independently associated with age, BP-medication and plasma levels of adiponectin when adjusted for SBP, which was also associated with distensibility, but only with age and plasma adiponectin when not adjusted for BP. In further regression analysis for distensibility stratified by hypertensive status, age (non-hypertensive: β = −0.562, P<0.0001; hypertensive: β = −0.465, P = 0.007) and plasma adiponectin (non-hypertensive: β = 0.527, P = 0.001; hypertensive: β = 0.518, P = 0.023) were independently associated with distensibility in non-hypertensive and hypertensive groups, in addition to SBP in the hypertensive group (β = −0.465, P = 0.007).

Table 3

Multivariable linear Regression Analyses.


The present study demonstrates that increased cIMT and reduced distensibility is observed in Class III obese subjects with no overt CVD conditions when compared to their age- and gender-matched non-obese counterparts. Changes in both cIMT and distensibility corresponded well with elevated traditional CVRF in this cohort. We also show that cIMT and distensibility are significantly associated with adiposity, adipokines and inflammatory markers, however, none had any connection with circulatory angiogenic cells.

Since obesity, and in particular abdominal obesity, is a major risk factor for CVD [23], tools to screen, monitor and predict CVD in this population can be very useful clinically. cIMT and distensibility, structural and functional parameters of carotid arteries, are early markers as well as strong predictors of CVD [1], [2]. While previous studies have demonstrated a strong association of increased cIMT and/or reduced distensibility with obesity [6][9], documentation of these two parameters of carotid artery in individuals with BMI>40 kg/m2 is lacking. Our finding that cIMT is increased and distensibility reduced in the Class III obese is consistent with the only other published studies that documented increased cIMT in 64 subjects with BMI of 42.3±4.3 kg/m2 [17], [18] and increased cIMT and decreased distensibility in 13 obese subjects with average BMI of 40.5±7 kg/m2 [18]. In addition, we show that cIMT is positively associated with age, adiposity, blood pressure, type-2 diabetes, hyperglycemia, dyslipidemia and hsCRP, while distensibility demonstrates negative associations with these covariates. These demonstrated associations verify that cIMT and distensiblity remain as CVD markers in Class III obesity.

Obesity is highly associated with the metabolic syndrome (MS), which is closely linked to cardiovascular morbidity and mortality [24]. Two of the most widely accepted MS criteria have been respectively promulgated by the World Health Organization (WHO) and the National Cholesterol Education Program (NCEP-ATP III). The principal distinction between the two is that that the NCEP-ATP III emphasizes CVD risks whereas the WHO focus on insulin resistance [25]. We used NCEP-ATP III criteria for this study. Based on the MS criteria defined by the NCEP-ATP III [25], about 86% of our Class III obese subjects had MS: dyslipidemia, hyperglycemia (diabetes), hypertension, or central obesity. This extremely high prevalence of MS was highly associated with BMI in our cohort as evidenced by the close association of BMI with various measures of MS status (Table 2). It is thus not surprising that BMI and/or plasma adiponectin was found to be independently linked to cIMT and distensibility, besides age which is a strong predictor of arterial remodeling [26], [27].

Hypertension and diabetes mellitus are common in obesity, which is also evident in this study. 60% and 56% of our Class III obese subjects were, respectively, hypertensive or diabetic in comparison to 20% and 7% in the non-obese group. Indeed in our obese subjects only 14% were free of both diabetes and hypertension. As both hypertension and diabetes are important in development of CVD, it is therefore difficult to apportion relative contributions to hypertension, diabetes or obesity per se. Further, artery stiffness and blood pressure are two closely related factors and distensibility is a derivative parameter of pulse pressure. Indeed, distensibility was revealed to independently associate with SBP and BP-medication besides age and plasma adiponectin when BP was adjusted in its multivariate regression analysis. The fact that an independent link between distensibility and plasma adiponectin stands regardless of the adjustment of BP demonstrates the essential role of adiposity in development of CVD suggesting the paramount importance of weight control in prevention and reduction of CVD. Obesity is also known to be characterized by a chronic systemic inflammatory state [28]. This was reflected by elevated plasma levels of CRP in our study. Plasma levels of CRP were strongly correlated with all important measures, cIMT, distensibility, BMI/waist:hip and as well as SBP, demonstrating a crucial role of inflammation in the development of obesity and obesity associated CVD. To further directly assess vascular inflammation and integrity, plasma endothelial microparticles (EMP) were examined and compared in randomly selected subpopulations including both genders from both groups. EMP are small membrane vesicles, 0.1–1 µm in diameter, which are released from the endothelium following endothelial cell activation or injury by a process of exocytotic budding of the plasma membrane [29] sometimes referred to as endothelial blebbing. Increased plasma EMP levels can be detected by FACS [30] and have been reported in various CVD conditions [29], [31]. In patients presenting a characterized endothelial dysfunction, levels of circulating EMP are inversely correlated with the amplitude of flow-mediated dilatation, independent of age and pressure [31], [32]. EMP has thus been suggested as a novel surrogate marker of endothelial injury, which precedes CVD, and hence a novel potential biomarker of CVD [20]. To our knowledge, this is the first report measuring EMPs in the Class III obese population and we show, for the first time, that circulating levels of EMP are significantly elevated in obesity.

In addition, we also examined plasma levels of angiogenic cells, an established cellular biomarker of CVD, and their associated cytokine SDF-1, a potent stimulator to mobilize angiogenic cells [33]. Decreased circulating levels of endothelial progenitor cells (EPC) have been used as an indicator of higher CVD risk [19]. We measured angiogenic cells in our study as CD133+/KDR+ peripheral blood mononuclear cells (PBMCs) by FACS. In comparison to EPC, commonly defined as CD34+/KDR+ PBMCs, CD133+/KDR+ PBMCs are earlier endothelial progenitors [34]. The role of EPC or angiogenic cells in obesity, and particularly in the Class III obese is largely unknown albeit there have been some studies which demonstrate low circulating levels of EPC in obesity with suggestions of the potential to predict CVD prevalence in this population [35], although contrary reports have also been observed [36]. In the current study, circulating levels of angiogenic cells were unchanged. This result might relate to the difference of angiogenic cells carrying different surface markers. We suggest, however, that the difference, at least partly, is the outcome of the contradicting processes occurring with inflammation suppressing the generation and mobilization of bone marrow-derived EPC [37] and highly activated adipogenesis promoting increased release of EPC from adipose tissue [38]. The latter is supported with the finding that SDF-1, secreted by adipose stromal cells [39], was significantly increased in the obese cohort. Intriguingly, the functional capacity of angiogenic cells measured as CFU-Hill colonies was significantly increased in the Class III obese group. The finding of a recent report by Hirschi et al. might explain this dichotomy. It reported that CFU-Hill colonies comprise primarily of monocytes and macrophages [40], and indeed what we are observing in the current study with increased CFU-Hill may simply reflect the activated inflammatory status in Class III obesity, in line with hsCRP. Our result of no correlation between cIMT or distensibility with angiogenic cells indicates that these cells may not be a cellular biomarker of CVD in Class III obesity.

It is a limitation of this study that only BMI or waist:hip ratio, but not direct measurement of body fat or body composition, was measured to reflect adiposity. Also neither insulin levels nor insulin resistance were evaluated. Further, this is a cross-sectional study, from which no casual relationship can be further explored. A future study in a cohort of subjects with BMI>40 kg/m2 with no overt CVD and also free of MS, esp. hypertension and diabetes mellitus, would be extremely useful to delineate the complexity of CVD pathophysiology in this population. Such individuals, however, are uncommon and thus difficult to identify.

In conclusion, we document in the current study that increased cIMT and reduced distensibility are present in Class III obesity. cIMT and distensibility correlate closely with traditional CVRF, adiposity and inflammatory markers, confirming the validity of these two important parameters in CVD detection in individuals with BMI>40 kg/m2. We also demonstrate, for the first time, elevated plasma EMP levels and unchanged circulatory CD133+/KDR+ angiogenic cells in Class III obesity.


We thank Ms Elizabeth Dewar and Ms Sofie Karapanagiotidis for their technical assistance on carotid imaging and measurement.

Funding Statement

This study was supported by a National Heart Foundation of Australia Project Grant (APP1012003) and in part by the Victorian Government’s Operational Infrastructure Support Program. MRS is a National Health and Medical Research Council of Australia (NHMRC) Career Development Fellow. AMD, JCD and JBD are NHMRC Fellows. Funding also provided by National Heart Foundation of Australia:, Victorian Government’s Operational Infrastructure Support Program: and National Health and Medical Research Council of Australia: The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.


1. Polak JF, Pencina MJ, Pencina KM, O’Donnell CJ, Wolf PA, et al. (2011) Carotid-wall intima-media thickness and cardiovascular events. N Engl J Med 365: 213–221. [PMC free article] [PubMed]
2. Blaha MJ, Budoff MJ, Rivera JJ, Katz R, O’Leary DH, et al. (2009) Relationship of carotid distensibility and thoracic aorta calcification: multi-ethnic study of atherosclerosis. Hypertension 54: 1408–1415. [PubMed]
3. Harloff A, Strecker C, Reinhard M, Kollum M, Handke M, et al. (2006) Combined measurement of carotid stiffness and intima-media thickness improves prediction of complex aortic plaques in patients with ischemic stroke. Stroke 37: 2708–2712. [PubMed]
4. Peralta CA, Adeney KL, Shlipak MG, Jacobs D Jr, Duprez D, et al. (2010) Structural and functional vascular alterations and incident hypertension in normotensive adults: the Multi-Ethnic Study of Atherosclerosis. Am J Epidemiol 171: 63–71. [PMC free article] [PubMed]
5. Tentolouris N, Liatis S, Moyssakis I, Tsapogas P, Psallas M, et al. (2003) Aortic distensibility is reduced in subjects with type 2 diabetes and cardiac autonomic neuropathy. Eur J Clin Invest 33: 1075–1083. [PubMed]
6. Burke GL, Bertoni AG, Shea S, Tracy R, Watson KE, et al. (2008) The impact of obesity on cardiovascular disease risk factors and subclinical vascular disease: the Multi-Ethnic Study of Atherosclerosis. Arch Intern Med 168: 928–935. [PMC free article] [PubMed]
7. Tounian P, Aggoun Y, Dubern B, Varille V, Guy-Grand B, et al. (2001) Presence of increased stiffness of the common carotid artery and endothelial dysfunction in severely obese children: a prospective study. Lancet 358: 1400–1404. [PubMed]
8. Recio-Rodriguez JI, Gomez-Marcos MA, Patino-Alonso MC, Agudo-Conde C, Rodriguez-Sanchez E, et al. (2012) Abdominal obesity vs general obesity for identifying arterial stiffness, subclinical atherosclerosis and wave reflection in healthy, diabetics and hypertensive. BMC Cardiovasc Disord 12: 3. [PMC free article] [PubMed]
9. Elkiran O, Yilmaz E, Koc M, Kamanli A, Ustundag B, et al. . (2011) The association between intima media thickness, central obesity and diastolic blood pressure in obese and owerweight children: A cross-sectional school-based study. Int J Cardiol [Epub ahead of print]. [PubMed]
10. Skilton MR, Sieveking DP, Harmer JA, Franklin J, Loughnan G, et al. (2008) The effects of obesity and non-pharmacological weight loss on vascular and ventricular function and structure. Diabetes Obes Metab 10: 874–884. [PubMed]
11. Rosendorff C (2007) Hypertension and coronary artery disease: a summary of the American Heart Association scientific statement. J Clin Hypertens (Greenwich) 9: 790–795. [PubMed]
12. Laakso M (2010) Cardiovascular disease in type 2 diabetes from population to man to mechanisms: the Kelly West Award Lecture 2008. Diabetes Care 33: 442–449. [PMC free article] [PubMed]
13. Loehr LR, Rosamond WD, Poole C, McNeill AM, Chang PP, et al. (2009) Association of multiple anthropometrics of overweight and obesity with incident heart failure: the Atherosclerosis Risk in Communities study. Circ Heart Fail 2: 18–24. [PMC free article] [PubMed]
14. Cheriyath P, Duan Y, Qian Z, Nambiar L, Liao D (2010) Obesity, physical activity and the development of metabolic syndrome: the Atherosclerosis Risk in Communities study. Eur J Cardiovasc Prev Rehabil 17: 309–313. [PubMed]
15. Sturm R (2007) Increases in morbid obesity in the USA: 2000–2005. Public Health 121: 492–496. [PMC free article] [PubMed]
16. Lavie CJ, Milani RV, Ventura HO (2009) Obesity and cardiovascular disease: risk factor, paradox, and impact of weight loss. J Am Coll Cardiol 53: 1925–1932. [PubMed]
17. Sturm W, Sandhofer A, Engl J, Laimer M, Molnar C, et al. (2009) Influence of visceral obesity and liver fat on vascular structure and function in obese subjects. Obesity (Silver Spring) 17: 1783–1788. [PubMed]
18. Ketel IJ, Stehouwer CD, Henry RM, Serne EH, Hompes P, et al. (2010) Greater arterial stiffness in polycystic ovary syndrome (PCOS) is an obesity–but not a PCOS-associated phenomenon. J Clin Endocrinol Metab 95: 4566–4575. [PubMed]
19. Sen S, McDonald SP, Coates PT, Bonder CS (2011) Endothelial progenitor cells: novel biomarker and promising cell therapy for cardiovascular disease. Clin Sci (Lond) 120: 263–283. [PubMed]
20. Nozaki T, Sugiyama S, Sugamura K, Ohba K, Matsuzawa Y, et al. (2010) Prognostic value of endothelial microparticles in patients with heart failure. Eur J Heart Fail 12: 1223–1228. [PubMed]
21. Liang YL, Teede H, Kotsopoulos D, Shiel L, Cameron JD, et al. (1998) Non-invasive measurements of arterial structure and function: repeatability, interrelationships and trial sample size. Clin Sci (Lond) 95: 669–679. [PubMed]
22. Montes M, Jaensson EA, Orozco AF, Lewis DE, Corry DB (2006) A general method for bead-enhanced quantitation by flow cytometry. J Immunol Methods 317: 45–55. [PMC free article] [PubMed]
23. Folsom AR, Kushi LH, Anderson KE, Mink PJ, Olson JE, et al. (2000) Associations of general and abdominal obesity with multiple health outcomes in older women: the Iowa Women’s Health Study. Arch Intern Med 160: 2117–2128. [PubMed]
24. Potenza MV, Mechanick JI (2009) The metabolic syndrome: definition, global impact, and pathophysiology. Nutr Clin Pract 24: 560–577. [PubMed]
25. Fonseca VA (2005) The metabolic syndrome, hyperlipidemia, and insulin resistance. Clin Cornerstone 7: 61–72. [PubMed]
26. Juonala M, Kahonen M, Laitinen T, Hutri-Kahonen N, Jokinen E, et al. (2008) Effect of age and sex on carotid intima-media thickness, elasticity and brachial endothelial function in healthy adults: the cardiovascular risk in Young Finns Study. Eur Heart J 29: 1198–1206. [PubMed]
27. Noon JP, Trischuk TC, Gaucher SA, Galante S, Scott RL (2008) The effect of age and gender on arterial stiffness in healthy Caucasian Canadians. J Clin Nurs 17: 2311–2317. [PubMed]
28. Clement K, Langin D (2007) Regulation of inflammation-related genes in human adipose tissue. J Intern Med 262: 422–430. [PubMed]
29. Chironi GN, Boulanger CM, Simon A, Dignat-George F, Freyssinet JM, et al. (2009) Endothelial microparticles in diseases. Cell Tissue Res 335: 143–151. [PubMed]
30. Enjeti AK, Lincz LF, Seldon M (2007) Detection and measurement of microparticles: an evolving research tool for vascular biology. Semin Thromb Hemost 33: 771–779. [PubMed]
31. Feng B, Chen Y, Luo Y, Chen M, Li X, et al. (2010) Circulating level of microparticles and their correlation with arterial elasticity and endothelium-dependent dilation in patients with type 2 diabetes mellitus. Atherosclerosis 208: 264–269. [PubMed]
32. Esposito K, Ciotola M, Schisano B, Gualdiero R, Sardelli L, et al. (2006) Endothelial microparticles correlate with endothelial dysfunction in obese women. J Clin Endocrinol Metab 91: 3676–3679. [PubMed]
33. De Falco E, Porcelli D, Torella AR, Straino S, Iachininoto MG, et al. (2004) SDF-1 involvement in endothelial phenotype and ischemia-induced recruitment of bone marrow progenitor cells. Blood 104: 3472–3482. [PubMed]
34. Rustemeyer P, Wittkowski W, Greve B, Stehling M (2007) Flow-cytometric identification, enumeration, purification, and expansion of CD133+ and VEGF-R2+ endothelial progenitor cells from peripheral blood. J Immunoassay Immunochem 28: 13–23. [PubMed]
35. Muller-Ehmsen J, Braun D, Schneider T, Pfister R, Worm N, et al. (2008) Decreased number of circulating progenitor cells in obesity: beneficial effects of weight reduction. Eur Heart J 29: 1560–1568. [PubMed]
36. Bellows CF, Zhan Y, Simmons PJ, Khalsa AS, Kolonin MG (2011) Influence of BMI on Level of Circulating Progenitor Cells. Obesity (Silver Spring) 19: 1722–1726. [PMC free article] [PubMed]
37. Grisar J, Aletaha D, Steiner C W, Kapral T, Steiner S, et al. (2005) Depletion of endothelial progenitor cells in the peripheral blood of patients with rheumatoid arthritis. Circulation 111: 204–211. [PubMed]
38. Martinez-Estrada OM, Munoz-Santos Y, Julve J, Reina M, Vilaro S (2005) Human adipose tissue as a source of Flk-1+ cells: new method of differentiation and expansion. Cardiovasc Res 65: 328–333. [PubMed]
39. Zhao BC, Zhao B, Han JG, Ma HC, Wang ZJ (2010) Adipose-derived stem cells promote gastric cancer cell growth, migration and invasion through SDF-1/CXCR4 axis. Hepatogastroenterology 57: 1382–1389. [PubMed]
40. Hirschi KK, Ingram DA, Yoder MC (2008) Assessing identity, phenotype, and fate of endothelial progenitor cells. Arterioscler Thromb Vasc Biol 28: 1584–1595. [PubMed]

Articles from PLoS ONE are provided here courtesy of Public Library of Science

Modulation of Intestinal Inflammation by Yeasts and Cell Wall Extracts: Strain Dependence and Unexpected Anti-Inflammatory Role of Glucan Fractions

Download PDF
Jagadeesh Bayry, Editor


Yeasts and their glycan components can have a beneficial or adverse effect on intestinal inflammation. Previous research has shown that the presence of Saccharomyces cerevisiae var. boulardii (Sb) reduces intestinal inflammation and colonization by Candida albicans. The aim of this study was to identify dietary yeasts, which have comparable effects to the anti-C. albicans and anti-inflammatory properties of Sb and to assess the capabilities of yeast cell wall components to modulate intestinal inflammation. Mice received a single oral challenge of C. albicans and were then given 1.5% dextran-sulphate-sodium (DSS) for 2 weeks followed by a 3-day restitution period. S. cerevisiae strains (Sb, Sc1 to Sc4), as well as mannoprotein (MP) and β-glucan crude fractions prepared from Sc2 and highly purified β-glucans prepared from C. albicans were used in this curative model, starting 3 days after C. albicans challenge. Mice were assessed for the clinical, histological and inflammatory responses related to DSS administration. Strain Sc1-1 gave the same level of protection against C. albicans as Sb when assessed by mortality, clinical scores, colonization levels, reduction of TNFα and increase in IL-10 transcription. When Sc1-1 was compared with the other S. cerevisiae strains, the preparation process had a strong influence on biological activity. Interestingly, some S. cerevisiae strains dramatically increased mortality and clinical scores. Strain Sc4 and MP fraction favoured C. albicans colonization and inflammation, whereas β-glucan fraction was protective against both. Surprisingly, purified β-glucans from C. albicans had the same protective effect. Thus, some yeasts appear to be strong modulators of intestinal inflammation. These effects are dependent on the strain, species, preparation process and cell wall fraction. It was striking that β-glucan fractions or pure β-glucans from C. albicans displayed the most potent anti-inflammatory effect in the DSS model.


Probiotics are a popular alternative to antibiotics [1]. The positive effects of probiotics on humans and animals result either from a direct nutritional effect or a health effect, with probiotics acting as bioregulators of the intestinal microflora and reinforcing the host’s natural defences [2].

Saccharomyces cerevisiae var. boulardii (Sb) is described as a biotherapeutic agent in the clinical literature and is reported to be efficacious in the prevention of antibiotic-associated diarrhoea and colitis in humans [3], [4]. Orally administered Sb demonstrated clinical and experimental effectiveness in gastrointestinal diseases through modulation of host cell signalling pathways implicated in the pro-inflammatory response such as IL-1β and TNF-α. Sb exerts a trophic effect that restores intestinal homeostasis and activates expression of peroxisome proliferator–activated receptor-gamma which protects against gut inflammation [5].

It has recently been reported that Sb decreases inflammation and intestinal colonization by C. albicans in a BALB/c mouse model of colitis induced by dextran-sulphate-sodium (DSS) [6]. Interestingly, parallel studies indicated that Sb reduces C. albicans adhesion to human intestinal cell lines and decreases pro-inflammatory cytokine mRNA levels in response to C. albicans infection [7], [8].

Sb is generally administered as a lyophilized powder [9] and its use as a food additive has only been reported in a number of cases such as in the fermentation of raw vegetable materials [10] and incorporation into commercial yoghurts [11].Taxonomic studies indicate that Sb should be considered as a Saccharomyces cerevisiae strain [12], [13]. This leads to the question “do other strains of S. cerevisiae also possess probiotic properties?” [14].

In the present study, low doses of DSS were administered to mice for 2 weeks to induce colonic inflammation and promote the establishment of C. albicans colonization, followed by a 3-day restitution period. Either S. cerevisiae strains or glycan fractions were then administered daily by oral gavage for 2 weeks, starting 3 days after the C. albicans challenge, in order to assess their curative effects on both colonic inflammation and acceleration of colonic epithelial restoration. Using the DSS mouse model each dietary yeast was found to have its own effect on colitis and C. albicans colonization. The impact of orally administered glycan fractions extracted from S. cerevisiae that may reverse the adverse effects of DSS and C. albicans, and the biological activity of soluble β-glucan isolated from the C. albicans cell wall, were then investigated in this DSS mouse model.


Comparison of the probiotic potential of S. cerevisiae var. boulardii and S. cerevisiae 1-1 strain

S. cerevisiae 1-1 strain (Sc1-1) was selected from our collection as having previously exhibited a probiotic effect. This yeast is prepared as an active dry yeast so that it can react quickly to its environment (Table 1). As Sc1-1 was comparable in vitro to Sb in terms of decreasing growth and germ tube formation by C. albicans (data not shown), the aim was to compare the Sc1-1 and Sb strains for their ability to reduce C. albicans colonization and intestinal inflammation.

Table 1

Yeast strains used in the investigation.

The model used is summarized in Fig. 1A and can be defined as “curative”. In this model, a single dose of C. albicans was administered to mice receiving DSS. Either the Sb or Sc1-1 strain was given 3 days later when C. albicans colonization was established. Low mortality was observed in mice that received DSS or DSS+C. albicans whereas none of the mice given the same regimen plus S. cerevisiae died (Fig. 1B).

Figure 1

Comparison of the probiotic potential of S. cerevisiae var. boulardii and S. cerevisiae 1-1 strain.

Concerning C. albicans colonization in mice that received DSS (Fig. 1C), the number of colony-forming units (CFUs) in stools gradually increased from the day of C. albicans administration. From S. cerevisiae administration on day 4 to the endpoint, a dramatic reduction in number of C. albicans CFUs was observed. No difference was observed between the two S. cerevisiae strains in their activity on C. albicans clearance (Fig. 1C).

The cumulated clinical and histological scores (Fig. 1D–E) were higher in mice that received DSS or DSS+C. albicans whereas they were reduced significantly by the administration of either Sb or Sc1-1 (Fig. 1D–E). Both strains were equally effective in reducing intestinal inflammation assessed by these parameters (Fig. 2).

Figure 2

Histological analysis of DSS-induced colitis in mice.

To analyze the possible mediators involved in the reduction of the inflammatory response to DSS and C. albicans in the colon following S. cerevisiae administration, we focused on levels of TNF-α and IL-10 mRNA as representative pro- and anti-inflammatory cytokines. As shown in Fig. 1F–G, administration of either Sb or Sc1-1 was associated with a significant reduction in TNF-α expression and increase in IL-10 production. No difference was observed between the two S. cerevisiae strains in their ability to redirect the inflammatory response.

Analysis of the anti-C. albicans and anti-inflammatory properties of other S. cerevisiae strains with different preparation processes

As both Sc1-1 and Sb strains displayed identical beneficial effects in this specific model the Sc1-1 strain was used as the reference strain for probiotic activity. In this part of the study, the influence of yeast preparation procedure and other yeast strains was investigated.

From an initial screening involving 10 strains or preparation procedures, five representative examples of strain activities were selected and compared to Sc1-1. These consisted of strains Sc1-2, Sc2, Sc3 and Sc4.

Sc1-1 showed important differences in mortality (Fig. 3A), ability to reduce C. albicans colonization (Fig. 3B) and a decrease in both histological and clinical scores (Fig. 3C–D). The Sc1-1 strain gave excellent results for all parameters, as did Sc3, which is used for its probiotic activities (Fig. 4) Interestingly, some of the beneficial effects induced by Sc1-1 were abolished when this strain was prepared as an instant dry yeast (Table 1). The most striking results were observed with the Sc4 strain, which did not display any particular effect on C. albicans colonization over 2 days compared to the other S. cerevisiae strains. However, the mice were extremely constipated, clinically inflamed and highly colonized with C. albicans starting from day 7. From this point on, we discontinued collecting faeces as the mice were becoming extremely ill. This yeast, which presents optimal growth at low temperatures, dramatically exacerbated both the clinical and histological scores and was associated with high mortality (around 80%). We also found that both C. albicans and Sc4 increased the staging of colitis in mice (Fig. 4). These high clinical and histological scores were associated with high numbers of C. albicans CFUs in different parts of the gut from moribund mice (data not shown).

Figure 3

Assessment of biological effect of S. cerevisiae 1-1 versus other S. cerevisiae strains.
Figure 4

Histological analysis of DSS-induced colitis in mice.

Finally, Sc2, another industrial strain used for the production of yeast proteins, displayed an “intermediate” behaviour with reduced beneficial effects and slight worsening of mortality and clinical scores.

Identification of the cell wall fractions supporting the beneficial and adverse effects on C. albicans colonization and inflammation

For this purpose, Sc2 was used as this strain can be easily induced to undergo autolysis in order to prepare cell wall extracts including mannoprotein (MP) and β-glucan fractions (Fig. 5).

Figure 5

Schematic diagram of glycan preparation from S. cerevisiae.

In contrast to living cells, none of the cell wall extracts induced mortality (Fig. 6A). However, the changes in body weight and clinical scores were worsened considerably by MP fractions while these parameters were ameliorated by β-glucan fractions (Fig. 6B, F). Notably, the MP fraction as well as the Sc2 strain induced an important loss in body weight up to day 9 that was correlated with its incapacity to control C. albicans colonization (Fig. 6C). By contrast, administration of β-glucan fraction maintained normal body weight, reduced inflammation scores and promoted C. albicans clearance (Fig. 6A–F). Although the body weight of mice receiving MP fractions started to increase from day 10, the clinical activity score was higher than that of mice receiving β-glucan fractions (Fig. 6E).

Figure 6

Effect of glycan fractions derived from S. cerevisiae on Candida DSS-treated mice.

Activity of the homologous C. albicans oligoglucoside fraction in the DSS mouse model

As the glucoprotein fraction from S. cerevisiae unexpectedly displayed a protective effect in this model, it was decided to investigate both the structure and biological activity of the β-glucan fraction from C. albicans. The harsh whole cell extraction procedure leads to fraction-1, known as yeast ghosts (Fig. 7). Fraction-1 (F1) was analyzed by fluorescence microscopy with various fluorescent probes specific for cell wall glycans in comparison to zymosan, which is widely used for β-glucan immunological studies (Fig. 8). Both zymosan and F1 were labelled with monoclonal antibody (mAb) 2G8 specific for β-1,3 glucans and WGA which binds to chitin (Fig. 8). In contrast to zymosan, which was stained with both Concanavalin A (ConA) and GNL, no mannose residue signals were observed for F1. Immunofluorescent staining with antibodies to β-mannose, liable to be synthesized by C. albicans, was also negative with F1 (data not shown). Thus, yeast ghosts have no mannose residues in their cell wall (Fig. 8).

Figure 7

Schematic diagram of β-glucan preparation from C. albicans.
Figure 8

Immunofluorescence staining of C. albicans ghost cells and zymosan with various fluorescent probes specific for yeast cell wall glycans.

MALDI-MS analysis of the soluble fraction derived from F1 (F2) established that it consisted of a highly polydisperse hexose polymer consisting of 3–27 hexose residues (Fig. 9A). According to its reactivity with anti β-1,3 glucans, this component was susceptible to zymolyase digestion which produced a set of small water soluble fragments (2–5 Glc), as demonstrated by thin-layer chromatography (data not shown) and MALDI-MS analyses (Fig. 9B). After purification of this fraction by reverse phase and adsorption chromatography, its structure was established by NMR as a mixture of β1,3-substituted glucan oligomers with free reducing ends. Furthermore, these oligomers were shown to be partially substituted by a random single β1,6 glucopyranose residue (Fig. 9C).

Figure 9

Structural analysis of glycan fraction.

Mortality, weight loss, clinical activity, histological score and C. albicans colonization

After structure characterization of the β-oligoglucoside fraction (F2) extracted from C. albicans, the biological activity of F2 was tested in the DSS mouse model. After C. albicans challenge, mice were given F2 (1 mg/day) from day3 up to the endpoint (Fig. 1A). F2 administration significantly prevented mouse mortality due to either DSS or DSS+C. albicans (Fig. 10A). In contrast to DSS and DSS+C. albicans mice, which developed severe colitis and lost body weight, F2 administration reversed the adverse effect of colitis and the mice showed a significant amelioration of both the clinical inflammation score and body weight (Fig. 10B–D). Histological examination of colon sections from mice receiving either DSS or DSS+C. albicans showed important colonic inflammation which was associated with mucosal cell loss, crypt damage, mucosal ulceration and accompanying submucosal oedema (Fig. 10C–D). F2 administration significantly reduced colonic inflammation due to either DSS or DSS+C. albicans (Fig. 10C–D). C. albicans colonization in DSS-treated mice showed a steady increase as assessed by the number of CFUs in faeces which was consistent with the high load of C. albicans recovered from the stomach, ileum and colon of this group of mice at the endpoint of the experiments (Fig. 10E–F).

Figure 10

Effect of β-oligoglucoside fractions derived from C. albicans on Candida DSS-treated mice.

In contrast to the higher numbers of C. albicans CFUs recovered from different compartments of the gut in DSS-treated mice, oral administration of F2 decreased the number of C. albicans CFUs recovered from stools and all gut segments of DSS-treated mice (Fig. 10E–F). Furthermore, the reduction in clinical and histological scores was consistent with these low numbers of C. albicans CFUs in different parts of the gut (Fig. 11 and Table 2).

Figure 11

Summary of the effects of S. cerevisiae strains or glycan fractions on Candida DSS-treated mice.
Table 2

Effect of different strains and cell wall extracts on C. albicans colonization and inflammation in the curative C. albicans DSS model.


Excessive use of antifungal agents has been implicated in the emergence of antifungal resistance in C. albicans and constitutes a serious clinical problem in hospitals by affecting the natural balance of the intestinal microflora in these individuals [15]. The non-pathogenic yeast, Sb, which is widely prescribed for the treatment of antibiotic-induced gastrointestinal disorders and Clostridium difficile-associated enteropathies, has been shown to be an alternative approach to counterbalance the equilibrium of the intestinal microflora and modulate the innate immune defence [16], [17]. Orally administered Sb was successful in both the treatment of inflammatory bowel disease (IBD) and the elimination of C. albicans colonization [7], [8], [18][21].

Recently, it has been shown that Sb decreases both C. albicans colonization and intestinal inflammation in a mouse model of DSS-induced colitis [6]. Following this study, S. cerevisiae strains, MP and β-glucan fractions were screened in a mouse model of DSS-induced colitis. As Sb is considered taxonomically to be a strain of S. cerevisiae [12], [13], strain Sc1-1 was compared to Sb in the DSS model. Sc1-1 is a gastro-resistant strain that reacts rapidly to its environment and is widely used in the food industry. Incidentally, it was observed that both Sc1-1 and Sb strains reduced C. albicans filamentation in vitro and C. albicans adhesion to plastic-plate wells (data not shown). In the present study, we did not chose a prophylactic but a curative model in which the animals develop colitis with histological features that are similar to those seen in patients with IBD before starting their treatment.

In this model low doses of DSS were used in order to establish C. albicans colonization, followed by S. cerevisiae or yeast extracts administration to assess their effects on the inflamed colon and colonic epithelium restitution.

Two weeks of DSS administration were scheduled to induce moderate colonic inflammation in mice, with low mortality rates. A recent study by Samonis et al. showed that mice receiving a high daily oral dose of C. albicans (around 108 CFU/day) for 2 weeks did not respond to Sb treatment [22]. In our model, a single inoculum of C. albicans was used and Candida colonization was maintained naturally in the mouse gastrointestinal tract by the DSS-induced colitis since a high C. albicans dose could dramatically hide the beneficial effect of Sb. In the present study, and similar to the Sb strain, Sc1-1 decreased both C. albicans colonization and intestinal inflammation in terms of clinical and histological score and mortality. Another notable finding was the acceleration of colonic epithelium restoration in mice treated with these dietary yeasts leading to the absence of submucosal oedema and epithelial erosion. Mechanistically, a recent report on intestinal inflammation showed that Sb secretes motogenic factors that enhance intestinal epithelial cell restitution [23].

Regarding the RT-PCR results, both Sb and Sc1-1 reduced the expression levels of pro-inflammatory cytokine TNF-α mRNA in the colonic mucosa with subsequent enhancement of IL-10 mRNA expression that inhibits intestinal injury [24]. Additionally, different pro-inflammatory cytokines were investigated in this set of experiments and were consistent with TNF-α expression. Further investigation is required to determine the role of Th17/Treg responses in different sets of experiments [25], [26]. A recent study in patients with IBD showed that Sb reduced TNF-α production and significantly inhibited T-cell proliferation induced by intestinal inflammation [19]. Generally, the biological activities of S. cerevisiae in gastrointestinal inflammatory conditions are mediated through modulation of host pro-inflammatory responses not only by the whole yeast, but also by secreted factors able to interfere with host signalling molecules that control inflammation at different levels such as NF-κB [27], [28]. Sb produces a soluble anti-inflammatory factor that inhibits NF-κB activation and attenuates pro-inflammatory signalling in host cells. In addition, Sb stimulates IL-10 secretion from intraepithelial lymphocytes infected by C. albicans and Escherichia coli [29]. As Sc1-1 was shown to be comparable to Sb and presents the same beneficial features against C. albicans and intestinal inflammation, Sc1-1 was considered as the reference strain in the DSS model. To assess if the observed anti-inflammatory properties were strain-dependent, other S. cerevisiae strains were selected deliberately for their high phenotypic diversity. The possible influence of yeast preparation process on anti-C. albicans activity was also studied. Surprisingly, some strains had a dramatic effect in the DSS mouse model and the process of yeast preparation also had an influence on the yeast’s biological properties [30]. Each strain selected in this study was well characterized in vitro in terms of cell growth, osmostress, fermentation, viability and metabolites. However, different factors could influence the biological activity of the strains when introduced by gavage in the DSS mouse model: (i) the resistance of the cell wall related to the yeast preparation process [30]; (ii) the viability of the strain in the stomach, ileum and colon; (iii) its interaction with the microflora and intestinal mucosa [31]; and (iv) its ability to produce soluble anti-inflammatory factors in the milieu triggering expression of mediators in the intestinal epithelium and cells of monocyte lineage present in the submucosae [27]. Altogether, each strain has its own unique properties and supports specific activities within the host. The in vitro findings, together with the results for all S. cerevisiae strains analyzed in this study, suggest that Sc1-1 has beneficial biological activities reversing all aspects of colitis, including histological damage, diarrhoea and mucosal levels of the pro-inflammatory mediator TNF-α.

The cell wall is an essential structural component of yeast cells playing a central role in the interaction of yeasts with their environment. Unfortunately, the biological activities of S. cerevisiae cell wall components are still unclear in terms of C. albicans colonization and intestinal inflammation. Two components (MP and β-glucans) produced industrially were explored in our experimental model. With MP fraction administration, C. albicans colonization was not consistent with intestinal inflammation parameters, suggesting that MP fractions have differential effects on C. albicans colonization and intestinal inflammation. In contrast to MP fractions, GP fraction administration decreased the number of C. albicans CFUs concomitantly to all intestinal inflammation parameters.

Both of these components are known to be potent immunological activators, but their mechanisms of action are different and controversial [32][34]. As an example, both MP and β-glucans act positively on tumour cells and several microbial infections [34], [35]. Conversely, administration of β-glucans derived from C. albicans has been shown to exacerbate arthritis in mice [36]. Structurally, MP have extensive N-and O-linked mannosylation which serve as ligands for galectin-3 (Gal-3), mannose receptor and DC-SIGN on macrophages and dendritic cells [37]. Different MP express β-Man epitopes, which have been identified as the principal ligand for Gal-3 [38]. In a previous study using the DSS model with C. albicans, Gal-3 knock-out mice were less affected by intestinal inflammation and C. albicans colonization than wild-type animals [39]. Recently, it was shown that C. glabrata deficient in β-Man was less virulent in DSS-treated mice as revealed by low clinical and histological scores and reduction of C. glabrata colonization [40]. β-glucans have affinities towards different receptors such as CD11b/CD18 [41], located on neutrophils, or Dectin-1 on macrophages [42]. This results in β-glucan activation of cytokine production and in turn activation of adaptive immunity. Thus, β-glucans attenuate the impact of colitis compared to MP [43].

As our results also showed a beneficial effect of β-glucans on inflammation/colonization, insoluble ghost yeast cells derived from C. albicans containing β-glucans were prepared and compared to zymosan which is widely used in β-glucan studies where many investigators refer to it as β-glucan [44], [45]. Zymosan stimulates the production and activity of pro-inflammatory cytokines [45]. Additionally, when chemically characterized zymosan containing only β-(1–3)-glucans was added to macrophage cells, the production of IL-10, reactive oxygen species (ROS) and TNF-α increased in a dose-dependent way [46]. Bonifazi et al. demonstrated the capacity of zymosan to activate both inflammatory and tolerogenic dendritic cells (DCs) leading to the triggering of both Th17 and Treg cells in vivo [47]. Our observations showed that zymosan contains both mannans and β-glucans exposed together on the cell wall surface in comparison to C. albicans ghosts that contain only β-glucans. This evidence prevented us from further studies on zymosan. Different observations showed that the biological activities of soluble β-glucans differ from those of cell-associated β-glucans [32], [48], [49]. Ishibashi et al. showed that insoluble cell wall β-glucans induced intensive inflammatory and immunomodulating activities compared to soluble β-glucans [49]. Following the β-glucan analysis, the chemical structure of the soluble β-glucan fraction derived from C. albicans ghosts was characterized and its biological activities were tested in the DSS mouse model. Interestingly, orally administered β-glucans from C. albicans decreased intestinal inflammation and C. albicans colonization.Several reports show that β-glucan enhances the immune response and improves the clearance of pathogenic bacteria in animal models [50][52]; this supports our findings that smaller oligoglucosides derived from C. albicans showed beneficial activities against C. albicans and these results are comparable to β-glucans derived from S. cerevisiae. However, it may also be hypothesized that these individual oligoglucosides could block receptors such as dectin-1 and CD11b/CD18 and prevent multivalent binding necessary for strong triggering of the inflammatory responses [53]. Besides the importance of yeast molecules sensing for immune response, a third player may also possibly act in the general interplay. This is the mouse microbiota. Oligosaccharides are well known prebiotics active on the intestinal flora [54], [55], and although such a role has not been investigated for C. albicans derived oligoglucosides it cannot be ruled out. Altogether, these results demonstrate that oligoglucosides behave differently from the original C. albicans whole yeast cells in the DSS mouse model.

In summary, Sc1-1 was found to be comparable to Sb and had beneficial biological activities against C. albicans and intestinal inflammation. Clinical trials are currently being conducted with Sc1-1 and promising results have been seen in patients with IBD. In the second part of this study, we focused on cell wall components involved in direct contact with the host and demonstrated that, in contrast to MP, β-glucan fractions from either S. cerevisiae or C. albicans have a more potent anti-inflammatory effect against colonic colitis induced by DSS in mice. In conclusion, this study generated some progress in deciphering the nature of the yeast molecular components differentially favouring inflammation and/or C. albicans clearance. Future studies will include experiments on oligosaccharide administration to mice in order to determine how these glycans stimulate the growth of beneficial bacteria in the gut and boost the immune system providing therapeutic perspectives for digestive disorders and life-threatening fungal infections of endogenous origin.

Materials and Methods

Yeast strains

The yeast strains used in this study are shown in Table 1.

Preparation of β-glucan fractions from yeasts

The composition in dry matter of spray-dried S. cerevisiae Sc2 cell wall fractions is shown in Table 3 and the preparation procedure for MP and β-glucan fractions from the cell wall of the same strain is shown in Fig. 1. The fractionation and digestion procedure for extraction of the β-glucan fraction from C. albicans is summarized in Fig. 2. Briefly, the cell pellet of C. albicans (50 g wet weight) was incubated twice in 200 ml of 1 M NaOH at 70°C for 30 min. After washing with distilled water, the supernatant was removed and the pellet was oxidized with 100 mM NaIO4 (Sigma-Aldrich, France) at room temperature for 24 h in the dark [56]. After completion of the reaction, excess periodate was destroyed by adding ethylene glycol. After washing several times with water, the pellet was reduced with 1 M NaBH4 (Sigma Aldrich, France) at room temperature. The reaction was terminated by lowering the pH to 5 by the addition of acetic acid. After washing several times with water, the insoluble fraction was then lyophilized to produce fraction-1. Fraction-1 was treated with zymolyase 20T (0.2 mg/mL, Immuno™; ICN Biomedicals Inc.) at 37°C for 3 h. Zymolyase inactivation was performed at 70°C for 5 min. After centrifugation, the supernatant was dialyzed against distilled water. The dialyzed solution was loaded onto a Sep-Pak C18 column (Alltech) equilibrated with 0.1% TFA (trifluoroacetic acid). Eluate-1 was evaporated and the resulting oligoglucosaccharides were dissolved in distilled water and further purified on a carbograph column (Alltech carbograph SPE column). Eluate-2 from the carbograph column was lyophilized to produce fraction-2 (F2). Fluorescence microscopy was performed to assess surface oligomannose expression on fraction-1 in comparison to zymosan (Sigma-Aldrich). Fraction-1 and zymosan suspensions deposited on slides were incubated with either monoclonal antibody (mAb) 2G8 specific for β-1,3 glucans [57], [58], or wheat germ agglutinin(WGA), which binds to chitin [59], or Concanavalin A (ConA) or Galanthus nivalis lectin (GNL) or DAPI, as described previously [6], [40]. For animal experimentation, fraction-2 was suspended in water and divided into 200 µL aliquots (each aliquot of 200 µL contained 1 mg of β-glucans).

Table 3

Composition in dry matter of spray-dried Sc2 cell wall fractions.

Structural analysis of β-glucans

NMR experiments were performed at 300 K using a Bruker AvanceII 900 MHz spectrometer equipped with a 5 mm triple-resonance cryoprobe. Prior to NMR spectroscopic analyses in deuterium, oligosaccharides were repeatedly exchanged in 2H2O (99.97% 2H, Euriso-top; Saint-Aubin, France) with intermediate freeze-drying and finally dissolved in 2H2O and transferred into Shigemi (Allison Park, USA) tubes. Chemical shifts (ppm) were calibrated taking the methyl group from internal acetone at δ1H 2.225 and δ13C 31.55 ppm. MALDI-TOF mass spectra were acquired on a Voyager Elite DE-STR mass spectrometer (Perspective Biosystems, Framingham, MA). Prior to analysis, samples were prepared by mixing 1 µL of oligosaccharide solution (1–5 pmol) with 1 µL of 2,5 dihydroxybenzoic acid matrix solution (10 mg/mL in CH3OH/H2O, 50[ratio]50, vol/vol) directly on the target. Between 50 and 100 scans were averaged for each spectrum.


Six- to 8-week-old female BALB/c mice were used. All mice were maintained by Charles River Laboratories (France). Four sets of experiments were performed independently and each experiment was divided into control groups (eight mice/cage), including assessment of the effect of DSS alone, and experimental groups (10 mice/cage).

Ethics statement

All mouse experiments were performed according to protocols approved by the Subcommittee on Research Animal Care of the Regional Hospital Centre of Lille, France, and in accordance with the European legal and institutional guidelines (86/609/CEE) for the care and use of laboratory animals.

Inoculum preparation and induction of colitis

Each animal was inoculated on day 1 by oral gavage with 200 µL of phosphate-buffered saline (PBS) containing 107 live C. albicans cells. Mice were given 1.5% DSS (MW 36–50 kDa; MP Biomedicals, LLC, Germany) in drinking water from day 1 to day 14 to induce intestinal inflammation. Three days after C. albicans oral challenge, mice were administered by oral gavage with a single-daily dose of either 107 lyophilized S. cerevisiae strains or 1 mg of β-glucan fraction for 2 weeks. Lyophilized S. cerevisiae strains were rehydrated for 30 min in PBS at 37°C before administering to the mice [60]. The presence of yeasts in the intestinal tract was followed daily by performing plate counts of faeces (approximately 0.1 g/sample) collected from each animal [39]. The faecal samples were suspended in 1 mL saline, ground in a glass tissue homogenizer and plated onto Candi-Select medium (Bio-Rad Laboratories, Marnes la Coquette, France). This chromogenic medium is designed for the isolation of yeasts from clinical specimens and is intended to differentiate medically important yeast species depending on the colour of the colonies [61]. Colonies of C. albicans were counted after 48 h incubation at 37°C. The results were noted as colony forming units (CFUs)/µg of faeces.

Presence of C. albicans colonization in the gastrointestinal tract

To check for C. albicans colonization, the animals were sacrificed and the gastrointestinal tract was removed and separated into the stomach, ileum and colon. The tissues were cut longitudinally. After removal of intestinal contents, the tissues were washed several times in PBS to minimize surface contamination from organisms present in the lumen [62]. Serial dilutions of homogenates were performed. The results were noted as C. albicans CFUs/mg of tissue.

Assessment of clinical parameters

The mortality rate of DSS-treated mice was determined daily and a colon biopsy was taken immediately after death for histological analysis. Total body weight was measured daily. The data are expressed as mean percent change from starting body weight. Daily clinical activity score ranging from 0 to 8 was calculated as described elsewhere [39], [63].

Determination of histological score

Rings of the transverse part of the colon were fixed overnight in 4% paraformaldehyde-acid and embedded in paraffin for histological analysis. Cross-sections (4 µm thick) were stained with haematoxylin-eosin (Sigma-Aldrich, France). Histological scores were evaluated by two independent, blinded investigators who observed two sections per mouse at magnifications of ×10 and ×100. The scores were determined in accordance with Siegmund et al. [63] and the sections were evaluated for the following two subscores: (i) a score for the presence and confluence of inflammatory cells, including neutrophils, in the lamina propria and submucosa or transmural extension; and (ii) a score for epithelial damage, focal lymphoepithelial lesions, mucosal erosion and/or ulceration and extension to the bowel wall. The two subscores were added together and the combined histological score ranged from 0 (no changes) to 6 (extensive cell infiltration and tissue damage).

Real-time mRNA quantification

Total RNA was isolated from colon samples using a NucleoSpin RNA II kit (Macherey-Nagel, France) following the manufacturer’s instructions, with 20–50 units of DNase I (RNase-free) at 37°C for 30 min to avoid contamination with genomic DNA. RNA quantification was performed by spectrophotometry (Nanodrop; Nyxor Biotech, France). Reverse transcription of mRNA was carried out in a final volume of 26 µL from 1 µg total RNA using 300 U M-MLV reverse transcriptase (Invitrogen, France) according to the manufacturer’s instructions with 500 ng oligo(dT) 12–18 and 50 U ribonuclease inhibitor (RNase-Out, Promega). PCR was performed using an ABI 7000 prism sequence detection system (Applied Biosystems, France) with SYBR green (Applied Biosystems, France). Amplification was carried out in a total volume of 25 µL containing 0.5 µL of each primer [6], [39] and 1 µL of cDNA prepared as described above. SYBR green dye intensity was analyzed using Abiprism 7000 SDS software (Applera Corp.). All results were normalized to the housekeeping gene β-actin.

Statistical analysis

Data are expressed as the mean ± SE of five mice in each group. All comparisons were analyzed by the Mann-Whitney U test. Statistical analyses were performed using the StatView™ 4.5 statistical program (SAS Institute Inc., Meylan, France). Differences were considered significant when the P value was <0.05.


The authors thank Nadine FRANÇOIS, Caroline DUBUQUOY, Emilie GANTIER, and Edmone ERDUAL for their excellent technical assistance and Val HOPWOOD for editing the manuscript.

Funding Statement

This work was funded by the program LEVACI issued from the French Government research plan FUI 5th AAP (DGE-Lille University contract number 082906131, European funds FEDER and local funds from the Région Nord-Pas de Calais, Lille Métropole Communauté Urbaine). LEVACI partners belong to research clusters Nutrition-Santé-Longévité and Végépolys. This work was also funded by the FP7 Health 260338 “ALLFUN” project “Fungi in the setting of inflammation, allergy and auto-immune diseases: translating basic science into clinical practices.” The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.


1. Goldin BR, Gorbach SL (2008) Clinical indications for probiotics: an overview. Clin Infect Dis 46 Suppl 2: S96–100discussion S144–151. [PubMed]
2. Shanahan F (2000) Probiotics and inflammatory bowel disease: is there a scientific rationale? Inflamm Bowel Dis 6: 107–115. [PubMed]
3. Surawicz CM, Elmer GW, Speelman P, McFarland LV, Chinn J, et al. (1989) Prevention of antibiotic-associated diarrhea by Saccharomyces boulardii: a prospective study. Gastroenterology 96: 981–988. [PubMed]
4. Guslandi M, Giollo P, Testoni PA (2003) A pilot trial of Saccharomyces boulardii in ulcerative colitis. Eur J Gastroenterol Hepatol 15: 697–698. [PubMed]
5. Buts JP, De Keyser N (2010) Transduction pathways regulating the trophic effects of Saccharomyces boulardii in rat intestinal mucosa. Scand J Gastroenterol 45: 175–185. [PubMed]
6. Jawhara S, Poulain D (2007) Saccharomyces boulardii decreases inflammation and intestinal colonization by Candida albicans in a mouse model of chemically-induced colitis. Med Mycol 45: 691–700. [PubMed]
7. Murzyn A, Krasowska A, Augustyniak D, Majkowska-Skrobek G, Lukaszewicz M, et al. (2010) The effect of Saccharomyces boulardii on Candida albicans-infected human intestinal cell lines Caco-2 and Intestin 407. FEMS Microbiol Lett 310: 17–23. [PubMed]
8. Murzyn A, Krasowska A, Stefanowicz P, Dziadkowiec D, Lukaszewicz M (2010) Capric acid secreted by S. boulardii inhibits C. albicans filamentous growth, adhesion and biofilm formation. PLoS One 5: e12050. [PMC free article] [PubMed]
9. Periti P, Tonelli F (2001) Preclinical and clinical pharmacology of biotherapeutic agents: Saccharomyces boulardii. J Chemother 13: 473–493. [PubMed]
10. Sindhu SC, Khetarpaul N (2002) Effect of probiotic fermentation on antinutrients and in vitro protein and starch digestibilities of indigenously developed RWGT food mixture. Nutr Health 16: 173–181. [PubMed]
11. Lourens-Hattingh A, Viljoen BC (2001) Growth and survival of a probiotic yeast in dairy products. Food Research International 34: 791–796.
12. Mitterdorfer G, Mayer HK, Kneifel W, Viernstein H (2002) Protein fingerprinting of Saccharomyces isolates with therapeutic relevance using one- and two-dimensional electrophoresis. Proteomics 2: 1532–1538. [PubMed]
13. van der Aa Kuhle A, Jespersen L (2003) The taxonomic position of Saccharomyces boulardii as evaluated by sequence analysis of the D1/D2 domain of 26S rDNA, the ITS1-5.8S rDNA-ITS2 region and the mitochondrial cytochrome-c oxidase II gene. Syst Appl Microbiol 26: 564–571. [PubMed]
14. Martins FS, Nardi RM, Arantes RM, Rosa CA, Neves MJ, et al. (2005) Screening of yeasts as probiotic based on capacities to colonize the gastrointestinal tract and to protect against enteropathogen challenge in mice. J Gen Appl Microbiol 51: 83–92. [PubMed]
15. Arnold HM, Micek ST, Shorr AF, Zilberberg MD, Labelle AJ, et al. (2010) Hospital resource utilization and costs of inappropriate treatment of candidemia. Pharmacotherapy 30: 361–368. [PubMed]
16. van Nispen tot Pannerden CM, Verbon A, Kuipers EJ (2011) Recurrent Clostridium difficile infection: what are the treatment options? Drugs 71: 853–868. [PubMed]
17. Czerucka D, Piche T, Rampal P (2007) Review article: yeast as probiotics – Saccharomyces boulardii. Aliment Pharmacol Ther 26: 767–778. [PubMed]
18. Guslandi M, Mezzi G, Sorghi M, Testoni PA (2000) Saccharomyces boulardii in maintenance treatment of Crohn’s disease. Dig Dis Sci 45: 1462–1464. [PubMed]
19. Thomas S, Metzke D, Schmitz J, Dorffel Y, Baumgart DC (2011) Anti-inflammatory effects of Saccharomyces boulardii mediated by myeloid dendritic cells from patients with Crohn’s disease and ulcerative colitis. Am J Physiol Gastrointest Liver Physiol . [PubMed]
20. Avalueva EB, Uspenskii Iu P, Tkachenko EI, Sitkin SI (2010) [Use of Saccharomyces boulardii in treating patients inflammatory bowel diseases (clinical trial)]. Eksp Klin Gastroenterol 103–111. [PubMed]
21. Dalmasso G, Cottrez F, Imbert V, Lagadec P, Peyron JF, et al. (2006) Saccharomyces boulardii inhibits inflammatory bowel disease by trapping T cells in mesenteric lymph nodes. Gastroenterology 131: 1812–1825. [PubMed]
22. Samonis G, Falagas ME, Lionakis S, Ntaoukakis M, Kofteridis DP, et al. (2011) Saccharomyces boulardii and Candida albicans experimental colonization of the murine gut. Med Mycol 49: 395–399. [PubMed]
23. Canonici A, Siret C, Pellegrino E, Pontier-Bres R, Pouyet L, et al. (2011) Saccharomyces boulardii improves intestinal cell restitution through activation of the alpha2beta1 integrin collagen receptor. PLoS One 6: e18427. [PMC free article] [PubMed]
24. Yanaba K, Yoshizaki A, Asano Y, Kadono T, Tedder TF, et al. (2011) IL-10-producing regulatory B10 cells inhibit intestinal injury in a mouse model. Am J Pathol 178: 735–743. [PMC free article] [PubMed]
25. Zelante T, De Luca A, Bonifazi P, Montagnoli C, Bozza S, et al. (2007) IL-23 and the Th17 pathway promote inflammation and impair antifungal immune resistance. Eur J Immunol 37: 2695–2706. [PubMed]
26. Cheng SC, van de Veerdonk FL, Lenardon M, Stoffels M, Plantinga T, et al. (2011) The dectin-1/inflammasome pathway is responsible for the induction of protective T-helper 17 responses that discriminate between yeasts and hyphae of Candida albicans. J Leukoc Biol 90: 357–366. [PMC free article] [PubMed]
27. Sougioultzis S, Simeonidis S, Bhaskar KR, Chen X, Anton PM, et al. (2006) Saccharomyces boulardii produces a soluble anti-inflammatory factor that inhibits NF-kappaB-mediated IL-8 gene expression. Biochem Biophys Res Commun 343: 69–76. [PubMed]
28. Reed KL, Fruin AB, Gower AC, Gonzales KD, Stucchi AF, et al. (2005) NF-kappaB activation precedes increases in mRNA encoding neurokinin-1 receptor, proinflammatory cytokines, and adhesion molecules in dextran sulfate sodium-induced colitis in rats. Dig Dis Sci 50: 2366–2378. [PubMed]
29. Fidan I, Kalkanci A, Yesilyurt E, Yalcin B, Erdal B, et al. (2009) Effects of Saccharomyces boulardii on cytokine secretion from intraepithelial lymphocytes infected by Escherichia coli and Candida albicans. Mycoses 52: 29–34. [PubMed]
30. Chung YH, Walker ND, McGinn SM, Beauchemin KA (2011) Differing effects of 2 active dried yeast (Saccharomyces cerevisiae) strains on ruminal acidosis and methane production in nonlactating dairy cows. J Dairy Sci 94: 2431–2439. [PubMed]
31. Buts JP, De Keyser N (2006) Effects of Saccharomyces boulardii on intestinal mucosa. Dig Dis Sci 51: 1485–1492. [PubMed]
32. Lee JN, Lee DY, Ji IH, Kim GE, Kim HN, et al. (2001) Purification of soluble beta-glucan with immune-enhancing activity from the cell wall of yeast. Biosci Biotechnol Biochem 65: 837–841. [PubMed]
33. Garner RE, Hudson JA (1996) Intravenous injection of Candida-derived mannan results in elevated tumor necrosis factor alpha levels in serum. Infect Immun 64: 4561–4566. [PMC free article] [PubMed]
34. Chen J, Seviour R (2007) Medicinal importance of fungal beta-(1→3), (1→6)-glucans. Mycol Res 111: 635–652. [PubMed]
35. Hashimoto K, Okawa Y, Suzuki K, Okura Y, Suzuki S, et al. (1983) Antitumor activity of acidic mannan fraction from bakers’ yeast. J Pharmacobiodyn 6: 668–676. [PubMed]
36. Hida S, Miura NN, Adachi Y, Ohno N (2007) Cell wall beta-glucan derived from Candida albicans acts as a trigger for autoimmune arthritis in SKG mice. Biol Pharm Bull 30: 1589–1592. [PubMed]
37. Poulain D, Jouault T (2004) Candida albicans cell wall glycans, host receptors and responses: elements for a decisive crosstalk. Curr Opin Microbiol 7: 342–349. [PubMed]
38. Fradin C, Poulain D, Jouault T (2000) beta-1,2-linked oligomannosides from Candida albicans bind to a 32-kilodalton macrophage membrane protein homologous to the mammalian lectin galectin-3. Infect Immun 68: 4391–4398. [PMC free article] [PubMed]
39. Jawhara S, Thuru X, Standaert-Vitse A, Jouault T, Mordon S, et al. (2008) Colonization of mice by Candida albicans is promoted by chemically induced colitis and augments inflammatory responses through galectin-3. J Infect Dis 197: 972–980. [PubMed]
40. Jawhara S, Mogensen E, Maggiotto F, Fradin C, Sarazin A, et al. (2012) A murine model of dextran sulfate sodium-induced colitis reveals Candida glabrata virulence and contribution of beta-Mannosyltransferases. J Biol ChemIn press. [PMC free article] [PubMed]
41. Thornton BP, Vetvicka V, Pitman M, Goldman RC, Ross GD (1996) Analysis of the sugar specificity and molecular location of the beta-glucan-binding lectin site of complement receptor type 3 (CD11b/CD18). J Immunol 156: 1235–1246. [PubMed]
42. Brown GD, Gordon S (2001) Immune recognition. A new receptor for beta-glucans. Nature 413: 36–37. [PubMed]
43. Sener G, Sert G, Ozer Sehirli A, Arbak S, Uslu B, et al. (2006) Pressure ulcer-induced oxidative organ injury is ameliorated by beta-glucan treatment in rats. Int Immunopharmacol 6: 724–732. [PubMed]
44. Goodridge HS, Simmons RM, Underhill DM (2007) Dectin-1 stimulation by Candida albicans yeast or zymosan triggers NFAT activation in macrophages and dendritic cells. J Immunol 178: 3107–3115. [PubMed]
45. Taylor PR, Tsoni SV, Willment JA, Dennehy KM, Rosas M, et al. (2007) Dectin-1 is required for beta-glucan recognition and control of fungal infection. Nat Immunol 8: 31–38. [PMC free article] [PubMed]
46. Saijo S, Fujikado N, Furuta T, Chung SH, Kotaki H, et al. (2007) Dectin-1 is required for host defense against Pneumocystis carinii but not against Candida albicans. Nat Immunol 8: 39–46. [PubMed]
47. Bonifazi P, Zelante T, D’Angelo C, De Luca A, Moretti S, et al. (2009) Balancing inflammation and tolerance in vivo through dendritic cells by the commensal Candida albicans. Mucosal Immunol 2: 362–374. [PubMed]
48. Driscoll M, Hansen R, Ding C, Cramer DE, Yan J (2009) Therapeutic potential of various beta-glucan sources in conjunction with anti-tumor monoclonal antibody in cancer therapy. Cancer Biol Ther 8: 218–225. [PubMed]
49. Ishibashi K, Miura NN, Adachi Y, Ogura N, Tamura H, et al. (2002) Relationship between the physical properties of Candida albicans cell well beta-glucan and activation of leukocytes in vitro. Int Immunopharmacol 2: 1109–1122. [PubMed]
50. Hetland G, Ohno N, Aaberge IS, Lovik M (2000) Protective effect of beta-glucan against systemic Streptococcus pneumoniae infection in mice. FEMS Immunol Med Microbiol 27: 111–116. [PubMed]
51. Li J, Li DF, Xing JJ, Cheng ZB, Lai CH (2006) Effects of beta-glucan extracted from Saccharomyces cerevisiae on growth performance, and immunological and somatotropic responses of pigs challenged with Escherichia coli lipopolysaccharide. J Anim Sci 84: 2374–2381. [PubMed]
52. Liang J, Melican D, Cafro L, Palace G, Fisette L, et al. (1998) Enhanced clearance of a multiple antibiotic resistant Staphylococcus aureus in rats treated with PGG-glucan is associated with increased leukocyte counts and increased neutrophil oxidative burst activity. Int J Immunopharmacol 20: 595–614. [PubMed]
53. Forsyth CB, Mathews HL (2002) Lymphocyte adhesion to Candida albicans. Infect Immun 70: 517–527. [PMC free article] [PubMed]
54. Rousseau V, Lepargneur JP, Roques C, Remaud-Simeon M, Paul F (2005) Prebiotic effects of oligosaccharides on selected vaginal lactobacilli and pathogenic microorganisms. Anaerobe 11: 145–153. [PubMed]
55. Buddington KK, Donahoo JB, Buddington RK (2002) Dietary oligofructose and inulin protect mice from enteric and systemic pathogens and tumor inducers. J Nutr 132: 472–477. [PubMed]
56. Gastebois A, Mouyna I, Simenel C, Clavaud C, Coddeville B, et al. (2010) Characterization of a new beta(1–3)-glucan branching activity of Aspergillus fumigatus. J Biol Chem 285: 2386–2396. [PMC free article] [PubMed]
57. Sendid B, Dotan N, Nseir S, Savaux C, Vandewalle P, et al. (2008) Antibodies against glucan, chitin, and Saccharomyces cerevisiae mannan as new biomarkers of Candida albicans infection that complement tests based on C. albicans mannan. Clin Vaccine Immunol 15: 1868–1877. [PMC free article] [PubMed]
58. Torosantucci A, Bromuro C, Chiani P, De Bernardis F, Berti F, et al. (2005) A novel glyco-conjugate vaccine against fungal pathogens. J Exp Med 202: 597–606. [PMC free article] [PubMed]
59. Hilenski LL, Naider F, Becker JM (1986) Polyoxin D inhibits colloidal gold-wheat germ agglutinin labelling of chitin in dimorphic forms of Candida albicans. J Gen Microbiol 132: 1441–1451. [PubMed]
60. Martins FS, Veloso LC, Arantes RM, Nicoli JR (2009) Effects of yeast probiotic formulation on viability, revival and protection against infection with Salmonella enterica ssp. enterica serovar Typhimurium in mice. Lett Appl Microbiol 49: 738–744. [PubMed]
61. Sendid B, Francois N, Standaert A, Dehecq E, Zerimech F, et al. (2007) Prospective evaluation of the new chromogenic medium CandiSelect 4 for differentiation and presumptive identification of the major pathogenic Candida species. J Med Microbiol 56: 495–499. [PubMed]
62. Edwards-Ingram L, Gitsham P, Burton N, Warhurst G, Clarke I, et al. (2007) Genotypic and physiological characterization of Saccharomyces boulardii, the probiotic strain of Saccharomyces cerevisiae. Appl Environ Microbiol 73: 2458–2467. [PMC free article] [PubMed]
63. Siegmund B, Rieder F, Albrich S, Wolf K, Bidlingmaier C, et al. (2001) Adenosine kinase inhibitor GP515 improves experimental colitis in mice. J Pharmacol Exp Ther 296: 99–105. [PubMed]
64. Gillum AM, Tsay EY, Kirsch DR (1984) Isolation of the Candida albicans gene for orotidine-5′-phosphate decarboxylase by complementation of S. cerevisiae ura3 and E. coli pyrF mutations. Mol Gen Genet 198: 179–182. [PubMed]

Articles from PLoS ONE are provided here courtesy of Public Library of Science