Streptococcus infantarius subsp. infantarius (Sii) has been identified as predominant lactic acid bacteria in spontaneously fermented dairy products (FDPs) in sub-Saharan Africa including Côte d'Ivoire. However, Sii belongs to the Streptococcus bovis/Streptococcus equinus complex (SBSEC). Most SBSEC members are assumed to be involved as opportunistic pathogens in serious diseases in both humans and animals. A population-based cross-sectional survey, including 385 participants was conducted in Korhogo, northern Côte d'Ivoire, to identify risk factors for Sii fecal carriage, including consumption of local FDPs. A structured questionnaire was used to gather participant's socio-demographic and economic characteristics, their relation to livestock and dietary habits. In addition, fresh stool and milk samples were collected. The identification of Sii was done using a SBSEC-specific PCR assay targeting 16S rRNA and groEL genes. The overall prevalence of SBSEC and Sii carriage was 23.2% (confidence interval CI 95% = 18.9–27.5) and 12.0% (CI 95% = 8.4–15.5) for stool, respectively. Prevalence of Sii was significantly higher in consumers of artisanal butter compared with non-consumers (57.1% vs 10.1%, odds ratio OR: 11.9, 95% CI: 3.9–36.6), as well as in persons handling livestock (OR = 3.9; 95% CI = 1.6–9.3) and livestock primary products (OR = 5.7; 95% CI = 2.3–14.3). The closer contact with livestock was a risk factor for Sii fecal carriage. Sii strains were isolated from fresh and fermented milk products with a prevalence of 30.4% and 45.4%, respectively. Analysis of Sii population structure through the SBSEC multi locus sequence typing assay revealed a close relationship across human and dairy isolates, possibly linked to a Kenyan human isolate. All these outcomes underline the interest of in-depth investigations on the ecology, potential reservoirs and pathways of contamination by Sii at the human-animal-environment interface in comparison to yet to be collected data from Europe, Asia and the Americas to further elucidate the various roles of Sii.
Citation: Gboko KDT, Traoré SG, Sanhoun AR, Kirioua J, Otaru N, Kurt F, et al. (2019) Risk factors for the carriage of Streptococcus infantarius subspecies infantarius isolated from African fermented dairy products. PLoS ONE 14(11): e0225452. https://doi.org/10.1371/journal.pone.0225452
Editor: Gulcin Avci, Hitit University, TURKEY
Received: March 5, 2019; Accepted: November 5, 2019; Published: November 27, 2019
Copyright: © 2019 Gboko et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the manuscript and its Supporting Information files as well as all sequence files are available from the pubmlst.org database.
Funding: This project was part of the Safe Dairy consortium (KDTG, SGT, ARS, JK, NZ, FK, FNJ, JI, DWMK, BK, PR, JH, LM, CJ, RN and BB) funded by ERAfrica/PASRES (https://www.erafrica.eu/) “New Ideas” project No ERAfrica_NI_021 (Swiss Partners NZ, FK, JI, JH, LM and CJ funded by the Swiss National Science Foundation (http://www.snf.ch) Project No IZEAZ0_154308). This study is partially based on work supported by the UBS Optimus Foundation (KDTG, SGT, ARS, JK; NO, FK, JI, DWMK, PR, JH, LM, CJ and BB; https://www.ubs.com/microsites/optimus-foundation/de/home.html) and the DELTAS Africa Initiative (KDTG, SGT, ARS, JK and BB; Afrique One-ASPIRE/DEL-15-008 http://afriqueoneaspire.org/). Afrique One-ASPIRE is funded by a consortium of donors including the African Academy of Sciences (AAS) Alliance for Accelerating Excellence in Science in Africa (AESA), the New Partnership for Africa’s Development Planning and Coordinating (NEPAD) Agency, the Wellcome Trust [107753/A/15/Z] and the UK government. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
The Streptococcus bovis/ Streptococcus equinus complex (SBSEC) is a heterogeneous group of bacteria within Lancefield group D streptococci . Based on genetic analyses, members of the SBSEC have been delineated into a main Streptococcus gallolyticus branch comprised of S. gallolyticus subsp. gallolyticus (Sgg), S. gallolyticus subsp. pasteurianus (Sgp), S. gallolyticus subsp. macedonicus (Sgm), the Streptococcus infantarius branch encompassing S. infantarius subsp. infantarius (Sii) and S. lutetiensis (intermediately called S. infantarius subsp. coli) and the single species S. equinus and S. alactolyticus [2, 3].
Members of the SBSEC are part of the gastrointestinal microbiota of humans and animals [1, 3, 4]. The fecal carriage rate of SBSEC in healthy adults ranges from 5% to 60% depending on isolation and identification techniques as well as geography [3, 5–7]. Members of the SBSEC are also described as opportunistic pathogens associated with various disease in humans and animals [7–13]. However, the major interest of these bacteria in human pathology, besides disease associations, is their controversial relationship with malignancies, mainly colorectal neoplasm [7, 12, 14–18]. Frequency of colorectal neoplasm varies widely from 6 to 67% among patients with SBSEC-related diseases. screening parameters, diagnostic tools and SBSEC species and subspecies seem to have a large impact on this widely varying frequency . Sgg has been described as more strongly associated with colorectal cancer (CRC) and infective endocarditis in contrast to Sgp and S. infantarius branch members. The later two seem more frequently involved in hepatobiliary disorders and non-colonic cancer [7, 11, 14, 17, 19, 20]. Nevertheless, a colonoscopy is recommended for CRC screening in any patient with a SBSEC-related disease such as infective endocarditis, independent of the involved species. This was mainly attributed to the potential association of all SBSEC members in malignancy development [18, 21, 22]. Data supporting this association as well as documented infective endocarditis cases are however minimal for the S. infantarius branch and its two subspecies Sii and S. lutetiensis [7, 18, 23]. In addition, various Sii strains have been isolated from food products, mainly fermented dairy products (FDPs). Sii has been identified as predominant lactic acid bacteria (LAB) at 108 bacteria per milliliter in spontaneously FDPs in various countries of sub-Saharan Africa, including in Côte d’Ivoire where 30–40% of FDPs are affected [24–28].
Comparative genomics of a Sii dairy isolate and the human-derived type strain revealed adaptation of the dairy isolate to the dairy environment similar to that of Streptococcus thermophilus . However, the pathogenic background of Sii and SBSEC requires a comparative approach of multiple Sii isolates of various sources including humans for a comprehensive risk assessment. Human-derived Sii isolates, particularly those of Africa origin, and corresponding prevalence data are so far limited to a single East-African hospital-based study . In West Africa, epidemiological data on human Sii isolates are lacking and knowledge is limited to dairy strains originating from Côte d’Ivoire and Mali [26, 28].
Population structure analysis by multi locus sequence typing (MLST) of Sii isolates indicated a division into three major clades comprised of (i) human blood isolates, (ii) East African dairy (EAD) isolates (iii) combination of West African dairy (WAD) and East African human (EAH) isolates . This therefore suggested genetic differences between EAD and EAH isolates. This therefore implies the difficulty to portrait disease associations of human isolates, e.g. CRC and hemorrhoids, to those of dairy lineages in East Africa [6, 30]. However, the shared WAD EAH Sii clade, the current inexistence of a West African dairy lineage and the close relationship of some West African Sii isolates to those isolated from human bacteremia cases warrants a better East-West comparison and an enhanced health risk assessment in West Africa . Therefore, we aimed in this study to identify risk factors for Sii fecal carriage corresponding with consumption of local FDPs. Thus, we hypothesized that the fecal carriage of Sii is higher in consumers of local FDPs. Given the existing Sii data for Mali and Côte d’Ivoire [26, 28], the dairy-producing area of Korhogo in Northern Côte d’Ivoire, close to the Malian border, would provide a suitable study area.
Materials and methods
Study site and population
A population-based cross-sectional survey was conducted between May and October 2014 in Korhogo, an agro-pastoral city in northern Côte d’Ivoire enrolling 595 households and 394 eligible participants. Before starting the main survey, preliminary interviews were performed in January 2014 with the head of the regional veterinary laboratory (LANADA), a representative of the Fulani community and milk sellers at the main market of the city. This allowed identifying the main local dairy products sold and the areas of high consumption. Seven neighborhoods were identified and selected for the present study: Cocody, Delafosse, Sozoribougou, Koko, Route Kapele, Banaforo and Ahoussabougou. In addition to these main sites, 31 small-scale dairy farms located in 10 hamlets, in a radius of 2 to 6 km of the city were also included for a possible comparison between outcomes from these areas and those from neighborhoods (Fig 1). A pilot survey was performed on a site outside those selected for the main study. In this pilot, the questionnaire was tested and stool samples were collected to improve procedures for data collection.
The illustrated map shows the selected neighborhoods (in blue capital font) and villages (in red) where the farms are located. Overview map of Côte d’Ivoire reprinted from the World Factbook.
The study population for the survey consisted of adults aged 18 and older, consumers and non-consumers of dairy products permanently resident in the city. Besides these inclusion criteria, participants had to give their prior consent to participate in the interview and provide a fecal sample to be eligible for the study. Subjects who did not meet the inclusion criteria, as well as those who had undergone gastrointestinal surgery (<6 months before the survey) or who had received chemotherapy or radiotherapy and the pregnant women were not eligible.
Field study sampling method of participants
Cluster design and field spotting.
Numeric maps with boundaries of selected neighborhoods were designed using random Global Positioning System (GPS) coordinates registered on-site, the detailed numeric map of the city and Google Earth version 7.0. The number of participants to be sampled per neighborhood was allocated proportionally to the size of their population. Several random starting points were then generated around the on site GPS coordinates using a random point generator (http://www.geomidpoint.com/random/). Clusters for sampling were designed from starting points surrounded by four main streets. Furthermore, a specific location (school, mosque) was defined in each neighborhood as landmark from which tracks leading to the main starting points in a selected neighborhood were identified. The application “GPX viewer by vectura” was used to run the sampling tracks. Once a starting point was accurately spotted, the boundaries of the corresponding cluster were identified on site.
Participants sampling procedure.
A multi-stage cluster sampling was conducted for participant’s selection. After selecting the clusters, a random selection of households was performed to be sampled along the boundaries of the main clusters in each block. The number of participants to be recruited per main starting point ranged from four to five depending on the population size and the number of clusters in each neighborhood. Thus, from the main starting house (which was not sampled), the next one anti-clockwise was sampled. Then, from the first household sampled, every third house was selected, until the number of interviews required was obtained or until the interviewer returned to the starting position. In cases where sampling of a household did not result in an interview (refusal case, non-eligible participants, empty house or non-inhabited place), the next household was directly selected and after that, every third house. The additional clusters were used when the required number of interviews was not reached after sampling of the main cluster in a block.
We assumed that one parcel or plot of land corresponded to one household. In cases where more than one household was found inside the same parcel, they were considered as a single household and one participant was randomly selected from all the residents if the parcel belonged to the same family. If different families resided in the same parcel, one out of the households was sampled randomly as following: from the entry of the plot of land, all households were numbered anti-clockwise. A number was selected on a single use random digit table (sequence of digits from 0 to 9, randomly arranged in a table). The household with the corresponding number was chosen for the survey. The number used on the random digit table was excluded from further use.
Afterwards, participant selection in each household was conducted. Only one participant was randomly selected per household according to the following procedure: the members of the household present at home at the time of the survey and meeting the eligibility criteria were ranked and numbered from the youngest to the oldest. Using as well a random digit table, a number was selected and the person with the corresponding number was selected for the survey.
Questionnaire survey and anthropometric measurements.
A structured questionnaire, based on that of Kenyan hospital-based study , was adapted to local customs, diet and habits. The questionnaire was used to identify potential factors associated with the fecal carriage of Sii. It covered information on demographic, socio-economic and behavioral determinants. This included smoking status, alcohol consumption, and physical activity as well as nutritional habits and the relation to livestock in daily live and work.
Assessed food items included mainly dairy products and food groups such as meats (red meat and poultry), fish, eggs, pulse, tea, coffee, fruits, vegetables, tubers, plantain, cereals and dietary supplements. Participants were asked for their usual food consumption frequency per food item. Frequencies were split into five categories: 1–2 days per month; 1–2 days per week; 3–5 days per week; 6–7 days per week and unknown frequency. For each food item, the average quantity of consumption was recorded. The portion size was estimated using household units (teaspoon, cup 1) and metric units (milliliter, kilogram).
Information on participants' health status were collected. This included reporting of diseases diagnosed over the past year by a health professional as well as self-reported clinical symptoms that they think they have suffered from during the 5–7 days preceding the survey. The questionnaire was administered in French but also in “Dioula”, a local West African language and "Senoufo", the local dialect. The height and weight of each participant were measured by using standard procedures .
Stool and milk sample collection.
Samples included stool from each participant and animal-derived milk from each household if available. For stool collection, a sterile screw-capped tube (Sarstedt, Sevelen, Switzerland) with integrated spoon was given to each participant after the interview. The the procedure for self-collection of a stool sample was thereby explained. The specimens were packed in stomacher bags and transported in cool boxes with ice packs to the regional laboratory. Fresh and fermented milk samples were also collected in households where available. Fermented milk samples were further differentiated into traditional FDP (tFDP) from the informal or household sector and industrial FDP. Milk samples were taken using sterile screw-capped tubes and transported in cool boxes with ice packs separated from those boxes used for stool samples. Upon arrival at the laboratory the same day, all samples were stored at -20°C until subsequent Sii isolation as previously done [24, 28]. This procedure is enabled by the ability of the targeted SBSEC bacteria, belonging to fecal streptococci, to survive several freeze-thaw cycles and the protective effects of carbohydrates, fat and proteins in milk and feces [32, 33]
Microbiological and molecular laboratory methods
Enumeration and selective isolation of SBSEC members from fecal and milk samples.
Mitis-salivarius (MS) agar medium with tellurite solution 1% (Becton Dickinson, Allschwil, Switzerland) was used for enumeration and isolation of presumptive SBSEC and Sii from fecal samples (n = 385) as previously described . Briefly: after thawing, an aliquot (about 100 mg) of each fecal sample was homogenized in 9 mL of sterile peptone saline solution (1% peptone, 0.85% NaCl) using a vortex. The suspension was serially diluted and 100 μL plated onto MS agar followed by incubation for 48 hours aerobically at 37°C. Out of the five plates seeded, three with desired black or blue colony types in pinpoint or slimy appearance and cfu counts <300 were selected for bacteria enumeration and isolation. One to three single colonies were picked per morphology type and streak purified.
The enumeration and isolation of presumptive members of the SBSEC from milk products (n = 52) was performed using M17 agar media (Biolife, Milan, Italy) as previously described . The isolation procedure using MS and M17 agar media therefore matched that of previous human stool and dairy product sampling, respectively, to ensure continued comparability of data [6, 24, 28].
Procedures for bacterial identification.
Preliminary identification was performed via assessment of the catalase activity of purified strains using hydrogen peroxide (3% H2O2) . Only catalase-negative isolates were further processed. DNA extraction for molecular biology tests was performed on purified single colonies using a short cell lysis in a Tris-EDTA triton-X-based buffer . Identification of presumptive SBSEC isolates was further carried out through SBSEC-specific PCR assays based on sequence analysis of 16S rRNA (primer set 16S-SBSEC-fw, 16S-SBSEC-3-fw, 16S-SBSEC-4-fw, 16S-SBSEC-5-fw and 16S-inf-rev) and groEL (primer set groEL-fw and groEL-rev) genes as previously described [28, 35, 36].
Determination of population structure of SBSEC members.
The population structure of SBSEC isolates was analyzed via the SBSEC MLST assay and deposited on pubmlst.org as previously described . Clonal complex (CC) calculations were performed in Phyloviz 2.0 and eBURSTv3 [37, 38] for groups of sequence types (STs) sharing 7 or more out of 10 alleles with at least one other member of this group . The complete set of strains including reference strains used for population structure calculations is described in S1 Table.
The questionnaire data were double-entered in Microsoft access 2010 and compared with EpiInfo version 3.5.3 (Center for Diseases Control and Prevention; Atlanta, United States of America). Data were analyzed using STATA version 12.1 (Stata Corporation, College Station, Texas 77845 United States of America). The household wealth index to assess participants' socioeconomic status was created using the principal components analysis (PCA) of 13 household assets as previously described . To determine the risk factors for Sii fecal carriage, logistic regression was performed. Variables with p ≤ 0.2 and those, which are biologically plausible, were selected for the multivariate regression model. The selected explanatory variables were then tested for multicollinearity and fitted into the final multivariable logistic regression model.
The study was approved by the national ethics committees of Côte d'Ivoire (N 018/MSLS/CNER-dkn), Eidgenössische Technische Hochschule (ETH) Zurich, Switzerland (EK2013-N-78) and Kantonale-Ethik-Kommission Zurich, Switzerland (StV-Nr47/14). The study was conducted in accordance with the Declaration of Helsinki . Informed written consent was obtained from all participants prior to enrollment in this study.
Study population characteristics
Out of the 355 participants initially interviewed in the neighborhoods, 39 failed to provide a stool sample (Fig 2). These participants were excluded and replaced. Overall, 394 subjects were randomly selected in the seven neighborhoods within 163 clusters. After curation of participants using the defined inclusion and exclusion criteria, a total of 385 participants (355 from neighborhoods and 30 farmers) were included in the analyses for basic demographic, economic and health characteristics of the study population (Table 1). The study population was young (median age of 35 years (inter quartile range = 27–48)), with 62.8% (242/385) females. More than one third of participants were overweight with a prevalence of 37.7%. Ulcer and hypertension were the second more frequent morbidities observed, with prevalence of 21.8% and 15.2%, respectively.
Consumption patterns of local milk products
Overall, 79.7% of the participants were regular consumers of local dairy products, which were produced exclusively from local cow milk. Most of consumers (84.7%) stated that they heat their milk prior to consumption, while 23.8% asserted that they consume milk directly after milking (farmers) or purchase (households). Regarding FDPs bought at market and homemade products, the proportion of consumers were 37.1% and 34.8%, respectively. Stratified by farm and urban households, significantly more people in farms consumed raw milk (63% vs 19%, p < 0.0001) and homemade FDPs (57% vs 32%, p = 0.008) compared with households in urban areas, where more people heated milk (88% vs 53%, p < 0.0001) before the consumption (Fig 3).
Sii prevalence and determinants
Sii fecal carriage.
The overall prevalence of SBSEC and Sii was 23. 2% (CI 95% = 18.9–27.5) and 12.0% (CI 95% = 8.4–15.5), respectively. Prevalence of Sii was significantly higher among subjects over 60 years of age and those born in Sahelian countries (Table 2). The fecal carriage rate of Sii was slightly higher in males and in the lowest socioeconomic quartile but the relationships were not statistically significant. Concerning the consumption of local dairy products, the overall fecal carriage rate of Sii in consumers of these foods was similar to that of non-consumers (12.1% vs 11.6%; OR: 1.0; 95% CI: 0.5–2.4; p = 0.92). Moreover, prevalence of Sii was higher in consumer of raw milk (19.0%) and homemade FDPs (16.3%) but the observed differences were not statistically different. However, the Sii fecal carriage rate was significantly higher in consumers of artisanal butter compared with non-consumers (57.1% vs 10.1%, OR: 11.9, 95% CI: 3.9–36.6).
Regarding residence area, more Sii fecal carriers were found among participants from farms compared with urban households (39.1% vs 9.9%; OR = 5.8; 95% CI: 2.3–14.6). Similarly, there were significantly more fecal carriers of Sii among livestock owners (OR = 2.2; 95% CI = 1.02–4.6), in subjects with daily routine with livestock (OR = 3.9; 95% CI = 1.6–9.3) and those with daily routine with livestock primary products (OR = 5.7; 95% CI = 2.3–14.3).
In the multivariate logistic regression analysis (Table 3), the direct livestock contact was strongly associated with Sii fecal carriage (OR = 6.2; 95% CI = 1.03–37.5). Subjects over 60 years of age had a higher risk for Sii carriage (OR = 1.6; 95% CI = 0.4–6.1), but this association was not statistically significant. Similarly, consumers of unboiled milk (OR = 1.5; 95% CI = 0.4–5), homemade FDPs (OR = 1.6; 95% CI = 0.6–4.1) and poultry with skin (OR = 2.1; 95% CI = 0.8–5.7) showed a higher risk for Sii fecal carriage without statistically significant differences observed. These overall calculations were derived from the analysis of 4450 isolates of fecal origin (S3 Table).
Sii strains were isolated from 52 milk samples, with an overall prevalence of 37.8% (CI 95% = 23.0–52.5). Prevalence of Sii in fermented and raw milk samples was 45.4% and 30.4% respectively. These rates were not statistically different. These overall calculations were derived from the analysis of 79 and 84 isolates of fermented and raw milk origin, respectively (S3 Table).
SBSEC population structure
A total of 53 SBSEC isolates collected during this study were processed by MLST. This included 50 Sii, one Sgg, one Sgp and one S. lutetiensis. In addition, MLST data was extracted from 37 SBSEC genomes available on GenBank to optimize the general SBSEC portfolio for comparisons. In combination with SBSEC STs derived from a parallel conducted farm-level study in the Korhogo area (Sanhoun et al., submitted for peer-review) and already deposited STs, this yielded a total of 271 SBSEC isolates in the current database generating 210 unique STs. The majority of isolates were identified as Sii and clustered in major CCs CC158, CC90, CC71, CC177 and CC161. This was in addition to a few two-ST complexes and multiple singletons (Fig 4 and S1 Fig).
Sii isolates and meta data are indicated concerning sample type and geographic classification (A), country of sample/isolate origin (B) and presence of lacS/lacZ dairy adaptation (C) and highlighting of new isolates and STs added in this study (D).
The majority of new isolates joined CC158 to yield a total of 96 isolates representing 68 STs. CC158 is comprised of a mix of Ivorian and Mali dairy isolates as well as Ivorian and Kenyan human isolates. The predicted founder ST158 is of Kenyan human origin. Further key nodes in CC158 are represented by Ivorian-derived ST137, ST131 and ST87 featuring a mix of dairy and human fecal sources. Particularly ST87 represents a key node for the majority of WAD Sii isolates (Fig 4). It is comprised of dairy isolates and STs originally described from the Abidjan area of Côte d’Ivoire and Mali as well as four novel STs and isolates obtained during this study. Dairy adaptation marker genes lacS and lacZ were detected in 5/15 STs connected to node ST87.
The connection between ST87 and ST158 via ST131 and ST137 as main intermittent nodes was facilitated by a series of STs representing Ivorian human-derived and Malian dairy-derived, dairy-adapted Sii isolates. The direct connection was comprised of four human-derived STs (ST120, ST126, ST137 and ST103), three dairy-derived STs (ST50, ST179 and ST3) and one ST shared between human and dairy sources. All dairy-derived STs also featured lacS/lacZ dairy adaptations with the exception of ST131-isolates. Out of the human-derived STs (ST120, ST126, ST137 and ST103), lacS/lacZ dairy adaptations were found in ST120, ST126, and ST137. The isolates yielding these STs were obtained from tFDP consumers of which one also consumed raw milk. ST137 furthermore represents a main node for a side branch for STs representing multiple dairy-adapted Sii STs (9/10 STs including ST137) derived from West African human and dairy sources. Among the six human-derived STs (ST125, ST127, ST133, ST137, ST138 and ST149), five also featured lacS/lacZ of which two STs were derived from tFDP consumers. The remaining three (ST125, ST138 and ST149) were obtained from participants not consuming tFDP or raw milk. Similarly, out of the total of 15 STs representing lacS/lacZ dairy-adapted variants obtained from human participants, nine STs were linked to tFDP consumers. This suggeststhe potential for transfer of Sii via tFDP and a possible reservoir of Sii in tFDP.
A total of 40 isolates were obtained as pairs originating either from human and dairy samples of the same household (n = 14), raw and FDPs from the same household (n = 6) or from the same human participant (n = 20) (S2 Table). Among dairy product isolates of participant AR007, two out of six isolates and STs (ST138/AR007A22.1 and ST139/AR007B22.1) featured closely related STs and shared dairy adaptation lacS/lacZ (S2 Table). In this study, closely related STs and isolates such as ST143/AR004A26.1 and ST152/AR004F6.1, both non-dairy-adapted, were isolated from dairy and fecal samples from participant AR004, respectively. In addition, the same dairy sample of participant AR004 comprised a dairy-adapted isolate ST138/AR004A26.3. This isolate was not obtained from fecal samples and only featured a distant relationship within CC158 to ST143 and ST152. Similarly, samples of participant AR022 featured dairy-adapted isolates of ST179 in FDPs. However, fecal samples returned only non-dairy-adapted isolates of ST172. Participant AR038 returned non-dairy-adapted isolates of ST86 and ST143 from dairy products whereas the fecal sample returned a distant non-dairy-adapted isolate of ST154. This suggests that a link between the two habitats may exist in some cases only.
In contrast, participant AR037 returned dairy-adapted ST138 and ST119 from dairy and fecal samples alike. Parallel existence of dairy-adapted and non-adapted STs in fecal samples of participants AR017, CL026, GKDT016, TKF022, TKF029 and dairy samples of participants AR004, AR016 and AR064 (S1 Table) suggests bidirectional exchange of strains between the dairy and the human habitat.
No new connections via MLST were observed for the Ivorian Sgp (ST153) S. lutetiensis (ST171) and Sgg (ST170) despite the addition of 37 GenBank-derived strains for Sgg (n = 13), Sgp (n = 1), S. lutetiensis (n = 3) and S. equinus (n = 20) (S1 Fig). Only for Sgg, the formation of a CC comprised of ST16, ST32, ST183 and ST205 of European infective endocarditis isolates as well as a CC comprised only of ST203 and ST204 of African primate nasal swabs was observed (S1 Fig).
This study aimed to fill the knowledge gaps regarding the epidemiology of Sii in West African regions. For this purpose, the prevalence and risk factors for Sii carriage were investigated, as well as the relationship of Sii strains of West and East African origin.
The SBSEC prevalence of 23.2% in our study resulted from a population-based cross-sectional survey. To the best of our knowledge, this is the first evaluation of SBSEC fecal carriage in the general population. However, the actual SBSEC prevalence might have been underestimated given that our assay was focused at isolating Sii. Sii is the predominant subspecies isolated in the present study, with a true prevalence of 12.0%. Nevertheless, similar SBSEC and Sii fecal carriage rates were observed in prior surveys [6, 41]. Particularly, the SBSEC prevalence is similar to that of the Kenyan hospital-based survey, which used a similar isolation and identification approach . It might thus provide good justification for an East-West comparison of SBSEC and Sii prevalence in humans to be related to a number of local factors, notably nutritional and environmental factors.
A relationship between the consumption of dairy products and Sii carriage was observed in this study. Indeed, Sii was significantly more prevalent in consumers of artisanal butter than in non-consumers. This food also called “naré” in Fulani language, is a dairy product obtained from milk fat (raw or boiled milk) after fermentation and using traditional extraction processes. In contrast, this association was not observed with the other local dairy products. The variation noted in the relation of dairy products with Sii carriage lead us to imagine the possibility that the fermentation level or process could influence this relationship.
In addition, MLST phylogeny suggested a link of Sii human isolates with dairy products. This was suggested by the presence of the dairy adaptation marker genes in some Sii isolates obtained from tFDPs consumers. This was not the case for EAH Sii isolates . More interestingly, although in small numbers, Sii human-dairy product strain pairs were also observed. This might suggest the contribution of tFDPs to the gut colonization by Sii, especially since Sii dairy strains may survive in the gastrointestinal tract. This finding in combination with the unexpected high proportion of Sii carriers in participants from Sahelian countries hints toward a potential contribution of cultural determinants, dietary habits and livestock culture to the gut colonization by these microorganisms. This should warrant investigations also in comparison to European or Asian or North American countries where comparable data is missing.
Interestingly, proximity to livestock seems to be a risk factor for Sii carriage. An analysis of the participants' relation to animals revealed a significantly higher proportion of Sii carriers only among those in direct contact with livestock and livestock primary products. Importantly, this closer livestock contact was identified as a risk factor in the multivariate analysis suggesting the possible transmission of Sii from animals to humans. This supports previous assumptions of hospital-based surveys, which ascribed the high frequency of SBSEC-related infections among subjects living in rural areas to the closer contact with animals . Similarly, a recent study involving Sgg identified the closer animal contact as well as usage of manure as risk factors for Sgg transmission to humans . In addition, Sgg was described as potential zoonotic bacteria due to shared STs between cattle and humans isolates . Epidemiological data regarding Sii from animal source are lacking. However, selected lineages of Sii strains were involved in human infections and thus, the possibility of zoonosis should be investigated also for Sii.
The high rate of Sii fecal carriers observed in elderly is consistent with previous reports noting an association between advanced age and occurrence of SBSEC-related infection [21, 43–45]. This fact seems to be related to the weakening of the body's immune defenses and frequency of morbidities in this age group . This therefore indicates that SBSEC members could act as opportunistic pathogens.
Sii strains have been isolated from FDPs with a prevalence similar to that previously reported in Côte d'Ivoire . Unexpectedly, Sii strains have also been isolated from fresh milk. This adds a potential new reservoir to FDPs as the so far exclusive Sii reservoir in the dairy niche . It remains to be investigated whether cross-contamination between FDPs and raw milk through the environment, workers or utensils was responsible for this finding. This highlights once again the interest for in-depth investigations on the ecology and potential reservoirs as well as pathways of contamination by Sii at the human-animal-environment interface.
Population structure analysis via MLST revealed the close relationship of West and East African human Sii and WAD Sii through the formation of a main CC158. The current ST- and isolate set led to a predicted founder of CC158 as ST158 of Kenyan human origin. While the link between East and West African isolates is well-supported by the presented data, the founder prediction and thus the possible evolution of Sii in Africa remains to be confirmed by in-depth analysis of additional isolates. Nevertheless, the presence of dairy adaptations offers some suggestions for their evolution: the Sii isolates in CC158 displayed presence and absence of dairy adaptations alike, as observed via marker genes lacS and lacZ. Isolates with STs neighboring ST137 featured such dairy adaptations mainly in a small branch derived from ST141 (adjacent to ST131), a large branch derived from ST137 and the STs connecting ST131 with ST137. This allowed for a possible separation of CC158 into these predominantly dairy-adapted branches and two main human-derived branches. The two predominantly human-associated branches thereby concentrated around ST131 (for West Africa) and ST158 (for East Africa). Except for the ST158 branch, no branch seemed to be exclusive to either human or dairy origin. Among the dairy-adapted branches, dairy products or the isolation from consumers of dairy products was the predominant isolate background. In the other branches and isolates, dairy adaptation was less prevalent and the isolation origin was less frequently associated with dairy consumers. This was particularly observed in STs linked to ST87 and its non-dairy-adapted isolates originating from dairy sources. This suggests for one the divergence of mainly dairy-adapted lineages from likely human commensal lineages. Second, bidirectional transfer of dairy-adapted strains to humans via FDP consumption and non-dairy-adapted human commensal strains to dairy products is a possible explanation for the presence of multiple STs of adapted and non-adapted isolates in the same habitat. The observation of multiple separate branches featuring dairy-adapted STs and the alternation between dairy-adapted and non-dairy-adapted STs in several branches strongly supports earlier indications for an evolution driven mostly by the occurrence of multiple recombination events rather than a single ancestral lineage. These findings would therefore support further investigations to better define the ecology and reservoirs of Sii in Africa. This would allow to better define marker strains for a safety assessment of Sii in relation to consumption via food.
This study provides the first data on SBSEC/Sii prevalence in the general population. Our findings allowed confirmation of earlier observations regarding determinants, especially demographic, environmental and nutritional factors that might influence the fecal carriage of SBSEC members. Such data was previously only obtained from hospital-based studies and was thus less representative for the general population. We noted a significantly higher rate of Sii fecal carriage in elderly, in consumers of artisanal butter, as well as in subjects in close contact with livestock and livestock primary products. The livestock proximity has been identified as a risk factor for Sii fecal carriage, indicating a possible transmission of Sii from animals to humans. The population structure of Sii in Côte d’Ivoire revealed a close relationship across human and dairy isolates in East and West Africa with a predicted founder linking back to a Kenyan human isolate. The presence of dairy adaptations and isolation origins however suggest the establishment of multiple dairy lineages and an evolution driven by recombination rather than clonal development. In-depth investigations on the ecology, potential reservoirs and pathways of contamination by Sii at the human-animal-environment interface in Africa will be needed in comparison to yet to be collected data from Europe, Asia and the Americas to further elucidate the various roles of Sii.
S1 Fig. Phylogenetic tree of SBSEC sequence types and isolates of dairy, plant and human origin.
The tree is based on MLST profiles and colored according to SBSEC species with additional indications for CC founders and dietary habits of the host.
S1 Table. SBSEC strains used in this study for population structure analysis.
All 271 SBSEC strains analyzed via MLST including relevant metadata are listed.
S2 Table. Key characteristics of the study participants featuring isolate pairs from fecal and dairy samples.
Study participants featuring isolate pairs from fecal and dairy samples with participant’s dietary habits and corresponding isolate information on sequence type (ST) and dairy adaption marker genes lacS/lacZ.
S3 Table. Complete list of fecal and milk isolates of the study.
All 4613 isolates processed in this study including identification results are listed.
The authors acknowledge the ERAfrica Safe Dairy consortium and particularly Wambui Kogi-Makau, Esther Schelling, Jakob Zinsstag, and Christophe Lacroix for their support to the overall project.
We would like to thank the interviewers and the Laboratoire National d'Appui au Développement Agricole, LANADA in Korhogo, Côte d’Ivoire for their support. We also thank Kaba Aboudramane, Dr Yapi Richard and Dr Aurelie Cailleau for their support in data management.
- 1. Herrera P, Min Kwon Y, Ricke SC. Ecology and pathogenicity of gastrointestinal Streptococcus bovis. Anaerobe. 2009;15:44–54. pmid:19100852
- 2. Schlegel L, Grimont F, Ageron E, Grimont PAD, Bouvet A. Reappraisal of the taxonomy of the Streptococcus bovis/Streptococcus equinus complex and related species: description of Streptococcus gallolyticus subsp. gallolyticus subsp. nov., S. gallolyticus subsp. macedonicus subsp. nov. and S. gallolyticus subsp. pasteurianus subsp. nov. Int J Syst Evol Microbiol. 2003;53:631–645. pmid:12807180
- 3. Jans C, Meile L, Lacroix C, Stevens MJA. Genomics, evolution, and molecular epidemiology of the Streptococcus bovis/Streptococcus equinus complex (SBSEC). Infect, Genet Evol. 2015;33:419–436. pmid:25233845
- 4. Schulz J, Dumke J, Hinse D, Dreier J, Habig C, Kemper N. Organic Turkey flocks: A reservoir of Streptococcus gallolyticus subspecies gallolyticus. PLoS One. 2015;10:e0144412. pmid:26657757
- 5. Dumke J, Vollmer T, Akkermann O, Knabbe C, Dreier J. Case-control study: Determination of potential risk factors for the colonization of healthy volunteers with Streptococcus gallolyticus subsp. gallolyticus. PLoS One. 2017;12:e0176515. pmid:28459818
- 6. Kaindi DWM, Kogi-Makau W, Lule GN, Kreikemeyer B, Renault P, Bonfoh B, et al. Investigating the association between African spontaneously fermented dairy products, faecal carriage of Streptococcus infantarius subsp. infantarius and colorectal adenocarcinoma in Kenya. Acta Trop. 2017;178:10–18. pmid:29079186
- 7. Jans C, Boleij A. The road to infection: Host-microbe interactions defining the pathogenicity of Streptococcus bovis/Streptococcus equinus complex members Front Microbiol. 2018;9:603. pmid:29692760
- 8. Grant RJ, Shang WY, Whitehead TR. Isolated septic arthritis due to Streptococcus bovis. Clin Infect Dis. 1997;24:1021. pmid:9142824
- 9. Fernández-Ruiz M, Villar-Silva J, Llenas-García J, Caurcel-Díaz L, Vila-Santos J, Sanz-Sanz F, et al. Streptococcus bovis bacteraemia revisited: clinical and microbiological correlates in a contemporary series of 59 patients. J Infect. 2010;61:307–313. pmid:20659498
- 10. Counihan-Edgar KL, Gill VA, Doroff AM, Burek KA, Miller WA, Shewmaker PL, et al. Genotypic characterization of Streptococcus infantarius subsp. coli isolates from sea otters with infective endocarditis and/or septicemia and from environmental mussel samples. J Clin Microbiol. 2012;50:4131–4133. pmid:23052307
- 11. Corredoira J, Alonso MP, García-Garrote F, García-Pais MJ, Coira A, Rabuñal R, et al. Streptococcus bovis group and biliary tract infections: an analysis of 51 cases. Clin Microbiol Infect. 2014;20:405–409. pmid:24033711
- 12. Alozie A, Köller K, Pose L, Raftis M, Steinhoff G, Westphal B, et al. Streptococcus bovis infectious endocarditis and occult gastrointestinal neoplasia: experience with 25 consecutive patients treated surgically. Gut Pathog. 2015;7:27. pmid:26473016
- 13. Alvarez A, Jia Y, Garcia CJ, Rosas-Blum ED, Boman D, Zuckerman MJ. Streptococcus bovis Bacteremia in Neonates in a Predominantly Hispanic Population. Front Pediatr. 2015;3:92. pmid:26579502
- 14. Ruoff KL, Miller SI, Garner CV, Ferraro MJ, Calderwood SB. Bacteremia with Streptococcus bovis and Streptococcus salivarius: clinical correlates of more accurate identification of isolates. J Clin Microbiol. 1989;27:305–308. pmid:2915024
- 15. Gold JS, Bayar S, Salem RR. Association of Streptococcus bovis bacteremia with colonic neoplasia and extracolonic malignancy. Arch Surg. 2004;139:760–765. pmid:15249410
- 16. Corredoira-Sánchez J, García-Garrote F, Rabuñal R, López-Roses L, García-País MJ, Castro E, et al. Association between bacteremia due to Streptococcus gallolyticus subsp. gallolyticus (Streptococcus bovis I) and colorectal neoplasia: a case-control study. Clin Infect Dis. 2012;55:491–496. cis434 [pii] pmid:22563018
- 17. Corredoira J, Grau I, Garcia-Rodriguez JF, Alonso-Garcia P, Garcia-Pais MJ, Rabuñal R, et al. The clinical epidemiology and malignancies associated with Streptococcus bovis biotypes in 506 cases of bloodstream infections. J Infect. 2015;71:317–325. pmid:25982024
- 18. Ben-Chetrit E, Wiener-Well Y, Kashat L, Yinnon AM, Assous MV. Streptococcus bovis new taxonomy: does subspecies distinction matter? Eur J Clin Microbiol Infect Dis. 2017;36:387–393. pmid:27796646
- 19. Krishnan S, Eslick GD. Streptococcus bovis infection and colorectal neoplasia: a meta-analysis. Colorectal Dis. 2014;16:672–680. pmid:24824513
- 20. Boleij A, van Gelder M, Swinkels DW, Tjalsma H. Clinical importance of Streptococcus gallolyticus infection among colorectal cancer patients: systematic review and meta-analysis. Clin Infect Dis. 2011;53:870–878. pmid:21960713
- 21. Lazarovitch T, Shango M, Levine M, Brusovansky R, Akins R, Hayakawa K, et al. The relationship between the new taxonomy of Streptococcus bovis and its clonality to colon cancer, endocarditis, and biliary disease. Infection. 2013;41:329–337. pmid:22886774
- 22. Habib G, Lancellotti P, Antunes MJ, Bongiorni MG, Casalta JP, Del Zotti F, et al. 2015 ESC Guidelines for the management of infective endocarditis: The Task Force for the Management of Infective Endocarditis of the European Society of Cardiology (ESC). Endorsed by: European Association for Cardio-Thoracic Surgery (EACTS), the European Association of Nuclear Medicine (EANM). European Heart Journal. 2015;36:3075–3128. pmid:26320109
- 23. Corredoira Sánchez J, García-Garrote F, García-País MJ, Fernández AT, Gonzalez-Juanatey C, Alonso García MP. Endocarditis caused by Streptococcus infantarius subsp. infantarius: a report of two cases. Int J Clin Pract. 2014;68:653–654. pmid:24750527
- 24. Jans C, Bugnard J, Njage PMK, Lacroix C, Meile L. Lactic acid bacteria diversity of African raw and fermented camel milk products reveals a highly competitive, potentially health-threatening predominant microflora. LWT-Food Sci Technol. 2012;47:371–379.
- 25. Jans C, Meile L, Kaindi DWM, Kogi-Makau W, Lamuka P, Renault P, et al. African fermented dairy products–overview of predominant technologically important microorganisms focusing on African Streptococcus infantarius variants and potential future applications for enhanced food safety and security. Int J Food Microbiol. 2017;250:27–36. pmid:28364623
- 26. Wullschleger S, Lacroix C, Bonfoh B, Sissoko-Thiam A, Hugenschmidt S, Romanens E, et al. Analysis of lactic acid bacteria communities and their seasonal variations in a spontaneously fermented dairy product (Malian fènè) by applying a cultivation/genotype-based binary model. Int Dairy J. 2013;29:28–35.
- 27. Abdelgadir W, Nielsen DS, Hamad S, Jakobsen M. A traditional Sudanese fermented camel's milk product, Gariss, as a habitat of Streptococcus infantarius subsp. infantarius. Int J Food Microbiol. 2008;127:215–219. pmid:18774196
- 28. Jans C, Kaindi DWM, Böck D, Njage PMK, Kouamé-Sina SM, Bonfoh B, et al. Prevalence and comparison of Streptococcus infantarius subsp. infantarius and Streptococcus gallolyticus subsp. macedonicus in raw and fermented dairy products from East and West Africa. Int J Food Microbiol. 2013;167:186–195. pmid:24131584
- 29. Jans C, Follador R, Hochstrasser M, Lacroix C, Meile L, Stevens MJA. Comparative genome analysis of Streptococcus infantarius subsp. infantarius CJ18, an African fermented camel milk isolate with adaptations to dairy environment. BMC Genomics. 2013;14:200. pmid:23521820
- 30. Kaindi DWM, Kogi-Makau W, Lule GN, Kreikemeyer B, Renault P, Bonfoh B, et al. Colorectal cancer-associated Streptococcus infantarius subsp. infantarius differ from a major dairy lineage providing evidence for pathogenic, pathobiont and food-grade lineages. Sci Rep. 2018;8:9181. pmid:29907746
WHO. Physical status: the use and interpretation of anthropometry. Report of a WHO expert committee. Geneva: World Health Organization, 1995 Contract No.: TRS No 854.
- 32. Gunnarsdóttir R, Müller K, Jensen PE, Jenssen PD, Villumsen A. Effect of long-term freezing and freeze-thaw cycles on indigenous and inoculated microorganisms in dewatered blackwater. Environ Sci Technol. 2012;46:12408–12416. pmid:23113759
- 33. Hubálek Z. Protectants used in the cryopreservation of microorganisms. Cryobiology. 2003;46:205–229. pmid:12818211
- 34. Goldenberger D, Perschil I, Ritzler M, Altwegg M. A simple universal DNA extraction procedure using SDS and proteinase K is compatible with direct PCR amplification. PCR Methods Appl. 1995;4:368–370. pmid:7580932
- 35. Jans C, de Wouters T, Bonfoh B, Lacroix C, Kaindi DWM, Anderegg J, et al. Phylogenetic, epidemiological and functional analyses of the Streptococcus bovis/Streptococcus equinus complex through an overarching MLST scheme. BMC Microbiol. 2016;16:117. pmid:27329036
- 36. Jans C, Lacroix C, Meile L. A novel multiplex PCR/RFLP assay for the identification of Streptococcus bovis/Streptococcus equinus complex members from dairy microbial communities based on the 16S rRNA gene. FEMS Microbiol Lett. 2012;326:144–150. pmid:22092382
- 37. Feil EJ, Li BC, Aanensen DM, Hanage WP, Spratt BG. eBURST: Inferring patterns of evolutionary descent among clusters of related bacterial genotypes from multilocus sequence typing data. J Bacteriol. 2004;186:1518–1530. pmid:14973027
- 38. Spratt BG, Hanage WP, Li B, Aanensen DM, Feil EJ. Displaying the relatedness among isolates of bacterial species—the eBURST approach. FEMS Microbiol Lett. 2004;241:129–134. pmid:15598523
O'Donnell O, van Doorslaer E, Wagstaff A, Lindelow M. Analyzing Health Equity Using Household Survey Data: The World Bank.
- 40. World Medical Association. World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects. JAMA. 2013;310:2191–2194. pmid:24141714
- 41. Lopes PG, Cantarelli VV, Agnes G, Costabeber AM, d'Azevedo PA. Novel real-time PCR assays using TaqMan minor groove binder probes for identification of fecal carriage of Streptococcus bovis/Streptococcus equinus complex from rectal swab specimens. J Clin Microbiol. 2014;52:974–976. pmid:24391203
- 42. Corredoira J, Alonso MP, Pita J, Alonso-Mesonero D. Association between rural residency, group D streptococcal endocarditis and colon cancer? Clin Microbiol Infect. 2008;14:190–190.
- 43. Pergola V, Di Salvo G, Habib G, Avierinos JF, Philip E, Vailloud JM, et al. Comparison of clinical and echocardiographic characteristics of Streptococcus bovis endocarditis with that caused by other pathogens. Am J Cardiol. 2001;88:871–875. [pii]. pmid:11676950
- 44. Giannitsioti E, Chirouze C, Bouvet A, Béguinot I, Delahaye F, Mainardi JL, et al. Characteristics and regional variations of group D streptococcal endocarditis in France. Clin Microbiol Infect. 2007;13:770–776. pmid:17501973
- 45. Duval X, Papastamopoulos V, Longuet P, Benoit C, Perronne C, Leport C, et al. Definite Streptococcus bovis endocarditis: characteristics in 20 patients. Clin Microbiol Infect. 2001;7:3–10. pmid:11284936