2459

Journal of Food Protection, Vol. 68, No. 11, 2005, Pages 2459±2464 Copyright ᮊ, International Association for Food Protection

Research Note Inability of Bacterial Strains rhamnosus HN001 and Bi®dobacterium lactis HN019 To Induce Human Platelet Aggregation In Vitro

J. S. ZHOU,1* K. J. RUTHERFURD,1,2 AND H. S. GILL1,2²

1Institute of Food, and Human Health (formerly Milk and Health Research Centre), Massey University, Palmerston North, New Zealand; Downloaded from http://meridian.allenpress.com/jfp/article-pdf/68/11/2459/1677293/0362-028x-68_11_2459.pdf by guest on 01 October 2021 and 2Fonterra Innovation (formerly New Zealand Dairy Research Institute), Palmerston North, New Zealand

MS 05-66: Received 15 February 2005/Accepted 11 July 2005

ABSTRACT

Platelet aggregation contributes to the pathogenesis of infective endocarditis, and aggregation of platelets induced by lactobacilli is thought to be an important contributory factor in the development and progression of Lactobacillus endocarditis. The main purpose of this study was to examine the effect of immunity-enhancing probiotic strains Lactobacillus rhamnosus HN001 and Bi®dobacterium lactis HN019 on the activation and aggregation of human blood platelets. Whole blood samples from healthy individuals were incubated in vitro with HN001 or HN019 and subsequently labeled with platelet-speci®c mono- clonal antibodies, ¯uorescein isothiocyanate±conjugated anti-CD41a (expressed on normal platelets), and phycoerythrin-strep- tavidin±conjugated anti-CD62p (expressed on activated platelets) before analysis by ¯ow cytometry. Platelet-rich plasma was used to assist the gating of the platelet cluster. ADP and epinephrine were used as the physiological platelet activation agonists. Platelet aggregation±inducing strain Streptococcus sanguis 133-79 was used as a positive control strain. The mean ¯uorescence intensity of phycoerythrin and the percentage of platelets expressing the CD62p marker were used to assess the degree of platelet activation. The percentage of CD62p-positive platelets and the light scatter pro®les of the agonist-activated platelets were used to identify the occurrence and degree of platelet aggregation. HN001 and HN019 had no effect on spontaneous platelet activation and aggregation; they also failed to exacerbate the platelet aggregation activity induced by ADP and epi- nephrine. Therefore, these test probiotic strains HN001 and HN019 are less likely to participate in the pathogenesis of infective endocarditis or other thrombotic disorders with regard to platelet aggregation factors.

The interaction between platelets and bloodborne bac- platelet aggregation in vitro (i.e., Aggϩ phenotype, such as teria is likely involved both in the pathophysiological mech- Streptococcus sanguis 133-79) consistently caused IE, with anisms of septicemia and disseminated intravascular coagu- signi®cantly more vegetation (the bacterial colonization of lation and in bacterial infective endocarditis (IE). IE is one a platelet-®brin clot on the endothelial surface of the heart), of the most frequently encountered problems associated with more gross lesions in major organs, and greater mortality opportunistic infections caused by , par- than observed after inoculation with AggϪ strains (i.e., ticularly lactobacilli (3, 8, 25). Platelet activation and aggre- strains incapable of inducing aggregation). Isolates of cer- gation are believed to contribute to the pathogenesis of IE, tain bacteria, including Lactobacillus rhamnosus, from pa- and aggregation of platelets by lactobacilli is thought to be tients with IE have been reported to uniformly induce ir- an important contributory factor in the progression of Lac- reversible platelet aggregation in vitro (13, 18). tobacillus-associated endocarditis (2, 9, 13, 17). Some lactobacilli can produce enzymes (e.g., glycosi- Various authors have reported that platelets in plasma dases and proteases) that function in the breakdown of hu- activate and aggregate in response to incubation with se- man glycoproteins and the synthesis and lysis of human lected gram-positive bacteria, including ®brin clots. These processes could aid the colonization and strains (15, 16, 21). For example, Herzberg and coworkers survival of bacteria and the initiation of endocarditis veg- (17) reported that intravenous inoculation of New Zealand etation (14, 24, 25). Most oral lactobacilli are commonly white rabbits with a bacterial strain capable of causing able to induce platelet aggregation in vitro and could thus potentially cause IE in vivo if they came into contact with * Author for correspondence. Present address: Division of Rheumatology, platelets in blood (2, 13, 27). Immunology and Allergy, Brigham and Women's Hospital, Harvard One hypothesis is that lactic acid bacteria from the gas- Medical School, 1 Jimmy Fund Way, Smith Building Room 636, Boston, trointestinal tract or oral cavity can be introduced into the MA 02115, USA. Tel: 617-525-1285; Fax: 617-525-1310; E-mail: bloodstream as the result of poor dental hygiene, gastroin- [email protected]. ² Present address: Primary Institutes Research Victoria, Department of testinal lesions, or surgical procedures and thus may con- Primary Industries, Werribee Centre, 600 Sneydes Road, Werribee, VIC tribute to the pathogenesis of IE and other pathological pro- 3030, Australia. cesses associated with platelet aggregation (4). Thus, a lack 2460 ZHOU ET AL. J. Food Prot., Vol. 68, No. 11

of platelet aggregation potential has been suggested as one cytes and therefore was used as a pan marker for platelets in this of the important criteria for selection of bacterial strains for study. Anti-CD62p reacts with the 140-kDa membrane glycopro- use as (7, 13, 19). Investigation of the ability of tein P-selectin (formerly known as platelet activation±dependent IE-associated enzymes to produce or elicit the aggregation granule membrane protein or GMP-140), which is stored in the of platelets has been recommended when the intrinsic prop- a-granules of platelets and the Weibel-Palade bodies of endothelial cells, and is rapidly transported to the plasma membrane upon erties of probiotic strains are being evaluated (12, 20, 23, activation; it was used as a marker for activated platelets in this 25). L. rhamnosus HN001 and Bi®dobacterium lactis study. HN019 are two newly identi®ed probiotic strains (26) with Tyrode's buffer (11) (1% bovine serum albumin, 2 mM/liter immune system±enhancing properties. For example, en- MgCl2, 137.5 mM/liter NaCl, 12 mM/liter NaHCO2, and 2.6 mM/ hancement of the phagocytic activity of granulocyte and liter KCl; pH 7.4), ACK lysis buffer (10), and a ®xative solution monocyte populations has been observed in the blood of (2% paraformaldehyde in PBS, pH 7.3) were prepared and stored human volunteers (5) and animals (10) following consump- at 4ЊC.

tion of B. lactis HN019 or/and L. rhamnosus HN001. In Downloaded from http://meridian.allenpress.com/jfp/article-pdf/68/11/2459/1677293/0362-028x-68_11_2459.pdf by guest on 01 October 2021 Collection and preparation of blood samples. Blood sam- mice fed a diet containing L. rhamnosus HN001 and B. ples were collected from six healthy human volunteers (24 to 45 lactis HN019, the number of B lymphocytes in the Peyer's years of age. Two samples from these donors and another blood patches in the gut increased and the proliferative responses sample from an additional donor were used in the validation ex- of spleen cells to mitogens was enhanced (10). In the cur- periment using S. sanguis 133-79. Written consent from these do- rent study, the ability of L. rhamnosus HN001 and B. lactis nors was obtained before blood collection. These individuals were HN019 to induce or enhance human blood platelet activa- nonsmokers, had consumed no alcohol within the previous 48 h, tion and aggregation in vitro was evaluated. and had not taken any drug during the previous 2 weeks. On the day of blood collection, a light standard breakfast was allowed MATERIALS AND METHODS before blood was collected (with the aid of a light tourniquet) from a forearm vein through a 19-gauge needle. Two milliliters Preparation of cultures. Stock cultures of test strains L. of blood was collected into a heparin-coated Vacutainer tube. rhamnosus HN001, B. lactis HN019, and a commercial probiotic Platelet-rich plasma was prepared by centrifugation of the whole strain L. rhamnosus GG (used as a reference strain) were obtained blood sample at room temperature (RT) at 300 ϫ g for 10 min. from Fonterra Innovation (formerly New Zealand Dairy Research Twenty-microliter samples of whole blood or platelet-rich plasma Institute) Culture Collection (Palmerston North, New Zealand). were immediately transferred into a set of 2-ml Eppendorf tubes These strains were propagated three times before being subjected containing Tyrode's buffer and different reagents or bacterial cells. to testing. Bacteria were grown in deMan Rogosa Sharpe (MRS; In these reaction systems, the ratio of bacterial cells to platelets BBL, Becton Dickinson, Sparks, Md.) broth for 18 h (i.e., to late was approximately 1:1. The samples were incubated at RT for 30 Њ stationary phase) at 37 C in an anaerobic jar (BBL Gaspack Sys- min, and then 100 ␮l of 2% formaldehyde (pH 7.3) was added to tem, Becton Dickinson) with Microbiology Anaerocult A (Merck, each tube to stop the reactions. After 5 min of incubation at RT, Darmstadt, Germany) and BBL Dry Anaerobic Indicator Strip 1 ml of ACK lysis buffer was added into each tube (whole blood (Becton Dickinson). For B. lactis HN019, MRS broth was sup- samples), which were incubated for another 10 min at RT to lyse plemented with 0.005% (wt/vol) cysteine hydrochloride (Sigma the red blood cells. The ®xed platelets were then harvested by Chemical Co., St. Louis, Mo.). Streptococcus sanguis 133-79 (Dr. centrifugation (MC 12V, DuPont Sorvall, Newtown, Conn.; 2,000 Mark Herzberg, School of Dentistry, University of Minnesota, ϫ g at RT for 10 min). Fifty microliters of Tyrode's buffer and 5 Minneapolis) is a prototype strain that induces platelet aggregation ␮l each of FITC-CD41a and PE-CD62p MoAbs or corresponding and was originally isolated from a patient with IE. S. sanguis 133- isotype controls (mouse IgG1, ␬) were added into each tube, Њ 79 cultures were grown overnight at 37 Cin5%CO2 in Todd- which were then incubated at RT in the dark for 20 min. After Hewitt broth (Remel, Lenexa, Kans.). The cells were harvested washing in Tyrode's buffer twice, the platelets were suspended in by centrifugation (Megafuge 1.0R, Heraeus Instruments, Germa- 200 ␮l of Tyrode's buffer and subjected to ¯ow cytometry anal- ϫ Њ ny; 3,000 g at 4 C for 15 min) and washed three times in cold ysis. phosphate-buffered saline (PBS, pH 7.3). The cells were resus- pended in PBS to an optical density of 0.5 Ϯ 0.02 at 610 nm and Flow cytometry analysis. Platelet samples treated as above then concentrated fourfold to give a suspension containing ap- were analyzed by ¯ow cytometry on a FACS Calibur (Becton proximately 109 CFU/ml (12, 19). Dickinson) within 30 min after labeling with MoAbs. Light scatter and ¯uorescence data were obtained with gain settings (FITC- Chemicals and monoclonal antibodies. ADP, epinephrine CD41a: 520 voltage; PE-CD62p: 566 voltage) in the logarithmic reagent (EPN), and prostaglandin E1 (PGE1) were purchased from mode (22, 29). The platelets were primarily distinguished from Sigma. ADP and EPN were used as physiological agonists to in- red blood cells and leukocytes on the basis of their forward- and duce platelet activation and aggregation, and PGE1 was used as side-light scatter pro®les and were further con®rmed by running an antagonist to block thrombin-induced platelet activation. Work- a platelet-rich plasma sample prepared following a standard meth- ing solutions of ADP (0.2 mM/liter), EPN (0.1 mM/liter), and od (6). Because of the much smaller scatter pro®les, debris or ␮ PGE1 (3 mM/liter) were prepared freshly before use. ``machine noise'' was excluded from the platelet gate position. Fluorescein isothiocyanate (FITC)±conjugated anti-CD41a Ten thousand cells within the platelet gate were analyzed, and the (mouse IgG1, ␬) and phycoerythrin (PE)-streptavidin±conjugated cells expressing FITC-CD41a (normal pan platelet marker) were anti-CD62p (mouse IgG1, ␬) monoclonal antibodies (MoAbs) selected (or back gated) for further analysis. The expression of were purchased from Pharmingen International (North Ryde, New PE-CD62p (the marker speci®c for activated platelets) was dis- South Wales, Australia). The anti-CD41a MoAb reacts with a cal- played in a quadrant dot plot (PE-CD62p ¯uorescence intensity cium-dependent complex of CD41-CD61 (GPIIb-IIIa) expressed versus FITC-CD41a ¯uorescence intensity). The mean ¯uores- on normal platelets (both resting and activated) and megakaryo- cence intensity (MFI) and percentage of PE-CD62p and FITC- J. Food Prot., Vol. 68, No. 11 PROBIOTIC BACTERIA AND PLATELET AGGREGATION 2461

TABLE 1. Platelet activation by probiotic strainsa Platelet treatment % CD62p-positive platelets (mean Ϯ SE) MFI index (mean Ϯ SE)

Ϯ Ϯ PGE1 5.185 0.9616 C 101.50 17.820 C HN019 8.864 Ϯ 0.9227 C 220.80 Ϯ 35.580 C HN001 4.658 Ϯ 0.8046 C 107.70 Ϯ 20.230 C GG 5.820 Ϯ 1.2516 C 145.50 Ϯ 37.740 C ADP ϩ ENP 64.25 Ϯ 4.2960 A 2,910.3 Ϯ 372.23 A S. sanguis 83.7 Ϯ 4.32 9,520 Ϯ 4,198 HN019 ϩ ADP ϩ ENP 64.27 Ϯ 11.060 A 2,596.3 Ϯ 935.12 AB HN001 ϩ ADP ϩ ENP 59.09 Ϯ 10.440 AB 2,249.5 Ϯ 748.64 AB GG ϩ ADP ϩ ENP 50.38 Ϯ 11.103 B 1,755.7 Ϯ 593.73 B a Fresh whole blood samples were incubated with test probiotic strains (HN001, HN019, or GG), the positive control strain (S. sanguis Downloaded from http://meridian.allenpress.com/jfp/article-pdf/68/11/2459/1677293/0362-028x-68_11_2459.pdf by guest on 01 October 2021 133-79), or ADP, epinephrine (EPN), or prostaglandin E1 (PGE1) for 30 min and then stained with platelet-speci®c antibodies CD41a (for normal platelets) and CD62p (for activated platelets). Only CD41a-positive cells were gated and analyzed. The percentage of CD62p-positive cells were gated and analyzed. The percentage of CD62p-positive platelets and the mean ¯uorescence intensity (MFI) index (MFI multiplied by the percentage of CD62p-positive cells) were calculated. Values are means Ϯ standard errors of a group of six samples (treated with probiotic strain and ADP, EPN, or PGE1) or three samples (treated with S. sanguis 133-79). Means followed by the same letter are not signi®cantly different (P Ͼ 0.001).

CD41a double-positive cells were used to evaluate the degree of nonlabeled and isotype control antibody-labeled platelets in plate- platelet activation or aggregation. The MFI index is the MFI mul- let-rich plasma and samples treated with ®xative or PGE1 before tiplied by the percentage of CD62p-positive cells. The antibody- labeling with PE-CD62p and FITC-CD41a. positive platelets were de®ned as those cells with ¯uorescence intensity higher than log 10, based on the ¯uorescence pro®les of Hematology. Total red blood cell and platelet counts, he- moglobin concentration, packed cell volume, mean corpuscular volume, mean corpuscular hemoglobin, and mean corpuscular he- moglobin concentration were determined on a Roche Cobas Mi- nos (ABX, Levallois, France).

RESULTS All hematology values were in the normal range. Platelet activation. There was no signi®cant difference in platelet activation between resting platelets and samples incubated with test probiotic strains. The platelets incubated with L. rhamnosus HN001, B. lactis HN019, or the refer- ence strain L. rhamnosus GG exhibited CD62p expression percentage and MFI index similar to those of resting plate- lets (4.6 to 8.8% and 107 to 220, respectively). However, in platelets treated with ADP and EPN there was a much higher percentage of CD62p-positive platelets (64.25%) and dramatically increased MFI index (2,910.3 Ϯ 372) compared with that of resting platelets (101 Ϯ 17.8) or probiotic-treated platelets (107 to 220 Ϯ 20 to 37) (Table 1 and Fig. 1). The activation of platelets induced by S. sanguis 133-79 is shown in Figure 2A and 2B. Similar to the effect of ADP and EPN, strain S. sanguis induced dra- FIGURE 1. Fluorescence pro®les of platelets incubated with test matic platelet activation under these experimental condi- bacterial strains. Fresh whole blood samples were treated and tions. More than 91% of CD41a-positive cells were CD62p stained as described in the ``Materials and Methods'' and the positive, and they had a very high MFI index (9,520 Ϯ footnote of Table 1. Only CD41a and CD62p double-positive cells 4,198). Incorporation of probiotic strains L. rhamnosus were selected for further analysis. The PE-CD62p ¯uorescence HN001, B. lactis HN019, or L. rhamnosus GG did not fur- intensity of these cells was visually displayed in a three-dimen- ther increase the expression of CD62p in samples treated sional histogram (cell count on y axis and PE-CD62p ¯uores- with ADP and EPN (Fig. 1). Incubation with test probiotic cence intensity on x axis). One representative plot from each strains actually caused a slight decrease in the percentage treatment group is shown. The peak at the left side of each plot indicates the ¯uorescence intensity of cells stained with isotype and MFI index of CD62p-positive platelets treated with ag- control antibodies, and the shaded peak at the right side of each onists compared with the samples treated with ADP and plot indicates the ¯uorescence intensity of cells stained with PE- EPN only, although the only signi®cant difference was for CD62p. samples treated with L. rhamnosus GG (Table 1). 2462 ZHOU ET AL. J. Food Prot., Vol. 68, No. 11

FIGURE 2. S. sanguis±induced platelet ac- tivation and aggregation. Fresh blood samples were treated with S. sanguis 133- 79 or 2% paraformaldehyde or ADP and epinephrine (EPN) as indicated in the foot- note of Table 1. CD41a and CD62p dou- ble-positive cells were selected and ana- lyzed. The ¯uorescence intensities (A and B) and light scatter pro®les (C) of these cells from one representative sample of each treatment are displayed. Downloaded from http://meridian.allenpress.com/jfp/article-pdf/68/11/2459/1677293/0362-028x-68_11_2459.pdf by guest on 01 October 2021

Platelet aggregation. Resting platelets and platelets vious adverse effects on the health of elder individuals fol- treated with test probiotic strains had a much lower forward lowing 3 weeks of consumption of B. lactis HN019 in low- scatter height than did platelets treated with ADP and EPN fat milk. In a more recent study, Tannock et al. (31) re- or S. sanguis 133-79. On the light scatter versus forward ported that consumption of L. rhamnosus HN001 for 6 height density plot, the cluster of platelets treated with ADP months is safe for humans. These observations indicate that and EPN (data not show) or with S. sanguis 133-79 (Fig. the probiotic strains tested in this study are unlikely to be 2C) was shifted up and to the right, indicating an increase pathogenic. in the size of platelet particles (aggregates). Blood samples However, bacteria from the gut or oral cavity can be incubated with probiotic strains had light scatter pro®les passively introduced into the blood stream by surgical or very similar to those of resting platelets. Incorporation of other invasive procedures; hence, these nonpathogenic bac- probiotics into the reaction systems of platelets treated with teria still may have the potential to cause infections or par- ADP and EPN did not change the light scatter pro®les of ticipate in the pathogenesis of some diseases, including IE. the activated platelets. These results demonstrate that L. Platelet aggregation is the key initiating event in the for- rhamnosus HN001, B. lactis HN019, and L. rhamnosus GG mation of a thrombus and makes an important contribution do not induce or exacerbate platelet aggregation. to the progression of IE. In light of these concerns, a bac- terial strain with platelet-aggregating properties may be DISCUSSION more virulent than nonaggregating strains. Thus, strains that L. rhamnosus HN001 and B. lactis HN019 were iso- do not aggregate platelets are likely to be safer than aggre- lated from cheese and , respectively, and have been gating strains for human consumption. consumed by humans for centuries with few safety con- L. rhamnosus GG was included as a reference strain cerns. Our earlier studies (32) revealed that L. rhamnosus because there are more safety data for this strain than for HN001 and B. lactis HN019 had no ability to degrade gas- any other commercial probiotic strain (27, 28) and because tric mucin. Arunachalam and coworkers (5) found no ob- it is unable to induce platelet aggregation in vitro (20). To J. Food Prot., Vol. 68, No. 11 PROBIOTIC BACTERIA AND PLATELET AGGREGATION 2463 assess the normal functions of tested platelets and observe acid bacteria with no platelet-aggregating activity might be any potentially exacerbatory effect of test probiotic strains limited in their ability to participate in the pathogenesis of on platelet aggregation under physiological conditions, IE. ADP and EPN were used as physiological triggers of plate- The probiotic strains L. rhamnosus HN001 and B. lac- let aggregation. Both PGE1 (an antagonist) and ®xatives tis HN019 are unable to induce spontaneous platelet acti- were used to achieve minimum activation of platelets that vation and do not enhance agonist-induced platelet aggre- were used as negative controls. An approximately 1:1 ratio gation processes. Therefore, they are safe probiotic strains of platelets to bacterial cells, which was reported as the in terms of their potential to participate in the pathogenesis optimum ratio for inducing platelet aggregation, was used of IE or other thrombotic disorders. in this study (13, 18, 21). S. sanguis 133-79, the strain that induces platelet aggregation (16), was included as a positive ACKNOWLEDGMENTS control to validate the experimental conditions and induced We thank Drs. Howard Katz and Frank Cross for their valuable sci- enti®c and editorial reviews, Dr. Mark Herzberg for his generous gift of dramatic platelet activation and aggregation. Flow cytom- Downloaded from http://meridian.allenpress.com/jfp/article-pdf/68/11/2459/1677293/0362-028x-68_11_2459.pdf by guest on 01 October 2021 etry using platelet-speci®c MoAbs is the most sensitive and S. sanguis 133-79, Fonterra Brands (formerly New Zealand Milk) for ®- nancial assistance, and all blood donors for their generous support. speci®c technique for studying platelet function (1, 6) be- cause of the advantages of using whole blood samples in- REFERENCES stead of platelet-rich plasma and avoiding the processes re- 1. Abrams, C. S., N. Ellison, A. Z. Budzynski, and S. J. Shattil. 1990. quired in conventional platelet aggregation methods to pre- Direct detection of activated platelets and platelet-derived micropar- pare samples (which may increase the chance of arti®cially ticles in humans. Blood 75:128±138. inducing platelet activation). Because interindividual vari- 2. Adams, M. R. 1999. Safety of industrial lactic acid bacteria. J. Bio- ation in the degree of augmentation of platelet aggregation technol. 68:171±178. 3. Adams, M. R., and P. Marteau. 1995. On the safety of lactic acid has been reported (13, 20), six whole blood samples were bacteria from food. Int. J. Food Microbiol. 27:263±264. included in this study to reduce the arti®cial effects caused 4. Aguirre, M., and M. D. Collins. 1993. Lactic acid bacteria and hu- by individual variation. In this study, no obvious interin- man clinical infection. J. Appl. Bacteriol. 75:95±107. dividual variation in the degree of augmentation of platelet 5. Arunachalam, K., H. S. Gill, and R. K. Chandra. 2000. Enhancement activation and aggregation was observed; these differences of natural immune function by dietary consumption of Bi®dobacter- ium lactis (HN019). Eur. J. Clin. Nutr. 54:263±267. in ®ndings may be related to the differences in experimental 6. Ault, K. A., and J. Mitchell. 1994. Analysis of platelets by ¯ow methods used in the current and previous studies. cytometry, p. 275±294. In Z. Darzynkiewicz, J. P. Robinson, and H. Under these experimental conditions, S. sanguis 133- A. Crissman (ed.), Methods in cell biology, vol. 42. Academic Press, 79 induced extensive platelet activation and aggregation San Diego. (Fig. 2) in all tested blood samples, but the test probiotic 7. Donohue, D. C., and S. Salminen. 1996. Safety of probiotic bacteria. Asia Pac. J. Clin. Nutr. 5:25±28. strains L. rhamnosus HN001, B. lactis HN019, and L. 8. Gallemore, G. H., R. T. Mohon, and D. A. Ferguson. 1995. Lacto- rhamnosus GG failed to induce or enhance any spontaneous bacillus fermentum endocarditis involving a native mitral valve. J. (or agonist-induced) platelet activation or aggregation. Be- Tenn. Med. Assoc. 8:306±308. cause L. rhamnosus HN001 and L. rhamnosus GG belong 9. Gasser, F. 1994. Safety of lactic acid bacteria and their occurrence to the same species, they have some characteristics in com- in human clinical infections. Bull. Pasteur Inst. 92:45±67. 10. Gill, H. S., K. J. Rutherfurd, J. Prasad, and P. K. Gopal. 2000. En- mon. According to Gasser (9) and Adams et al. (3), the hancement of natural and acquired immunity by Lactobacillus rham- overwhelming majority of reported endocarditis cases nosus (HN001), Lactobacillus acidophilus (HN017) and Bi®dobac- caused by nonenterococcal lactic acid bacteria have been terium lactis (HN019). Br. J. Nutr. 83:167±176. caused by Lactobacillus species, particularly L. rhamnosus 11. Ginsberg, M. H., A. L. Frelinger, S. C. T. Lam, J. Forsyth, R. Mc- and or . How- Millan, E. F. Plow, and S. J. Shattil. 1990. Analysis of platelet ag- gregation disorders based on ¯ow cytometric analysis of membrane ever, the results obtained in this study demonstrated that L. glycoprotein IIb-IIIa with conformation-speci®c monoclonal anti- rhamnosus HN001 is unlikely to be involved in the path- bodies. Blood 76:2017±2023. ogenesis of IE because of its inability to induce platelet 12. Hammes, W. P., N. Weiss, and W. Holzapf¯. 1992. The genera Lac- aggregation. In a total of 53 cases of endocarditis reported tobacillus and Carnobacterium, p. 1535±1594. In A. Balows, H. G. over 55 years (1938 to 1993), no cases were associated with Truper, M. Dworkin, W. Harder, and K. H. Schleifer (ed.), The pro- karyotes, vol. 2. Springer-Verlag, New York. Bi®dobacterium spp. (9). Therefore, there are no reports 13. Harty, D. W. S., M. Patrikakis, E. B. H. Hume, H. J. Oakey, and K. documenting the platelet aggregation properties of bi®do- W. Knox. 1993. The aggregation of human platelets by Lactobacillus bacteria. species. J. Gen. Microbiol. 139:2945±2951. According to Kirjavainen et al. (19), the incidence of 14. Herzberg, M. C. 1996. Platelet-streptococcal interactions in endo- bacteremia caused by lactic acid bacteria is quite low carditis. Crit. Rev. Oral Biol. Med. 7:222±236. 15. Herzberg, M. C., and K. L. Brintzenhofe. 1983. ADP-like platelet (0.24%, 8 of 3,317 blood culture isolates), and in most cas- aggregation activity generated by Viridans streptococci incubated es patients with cultures positive for lactic acid bacteria had with exogenous ATP. Infect. Immun. 40:120±125. a severe underlying disease that predisposed them to bac- 16. Herzberg, M. C., K. L. Brintzenhofe, and C. C. Clawson. 1983. teremic complications. Lactobacillus only rarely causes en- Aggregation of human platelets and adhesion of Streptococcus san- docarditis (30), with a reported rate of less than one case guis. Infect. Immun. 39:1457±1469. 17. Herzberg, M. C., G. MacFarlane, K. Gong, N. N. Armstrong, A. per year (9). Nevertheless, the ability to aggregate human Witt, P. R. Erickson, and M. W. Meyer. 1992. The platelet interac- platelets is a common attribute of the lactobacilli in the oral tivity phenotype of Streptococcus sanguis in¯uences the course of cavity of normal humans (13). Therefore, strains of lactic experimental endocarditis. Infect. Immun. 60:4809±4818. 2464 ZHOU ET AL. J. Food Prot., Vol. 68, No. 11

18. Kessler, C. M., E. Nussbaum, and C. U. Tuazon. 1987. In vitro cor- properties of lactic acid bacteria: set-up of an experimental model. relation of platelet aggregation with occurrence of disseminated in- IDF Nutr. Newsl. 5:24. travascular coagulation and subacute bacterial endocarditis. J. Lab. 26. Prasad, J., H. S. Gill, J. Smart, and P. K. Gopal. 1999. Selection and Clin. Med. 109:647±652. characterisation of Lactobacillus and Bi®dobacterium strains for use 19. Kirjavainen, P. V., E. M. Tuomola, R. G. Crittenden, A. C. Ouwe- as probiotics. Int. Dairy J. 8:993±1002. hand, D. W. S. Harty, L. F. Moris, H. Rautelin, M. J. Playne, D. C. 27. Salminen, S., V. A. Wright, L. Morelli, P. Marteau, D. Brassart, W. M. de Vos, R. Fonde'n, M. Saxelin, K. Collins, G. Mogensen, S. E. Donohue, and S. J. Salminen. 1999. In vitro adhesion and platelet Birkeland, and T. M. Sandholm. 1998. Demonstration of safety of aggregation properties of bacteremia-associated lactobacilli. Infect. probioticsÐa review. Int. J. Food Microbiol. 44:93±106. Immun. 67:2653±2655. 28. Saxelin, M. 1997. Lactobacillus GGÐa human probiotic strain with 20. Korpela, R., E. Moilanen, M. Saxelin, and H. Vapaatalo. 1997. Lac- thorough clinical documentation. Food Rev. Int. 13:293±313. tobacillus rhamnosus GG (ATCC 53103) and platelet aggregation in 29. Shattil, S. J., M. Cunningham, and J. A. Hoxie. 1987. Detection of vitro. Int. J. Food Microbiol. 37:83±86. activated platelets in whole blood using activation-dependent mono- 21. Kurpiewski, G., L. J. Forrester, B. J. Campbell, and J. T. Barrett. clonal antibodies and ¯ow cytometry. Blood 70:307±315. 1983. Platelet aggregation by Streptococcus pyogenes. Infect. Im- 30. Sussman, J. I., E. J. Baron, S. M. Goldberg, M. H. Kaplan, and R. A. Pizzarello. 1986. Clinical manifestations and therapy of Lacto- mun. 39:704±708. Downloaded from http://meridian.allenpress.com/jfp/article-pdf/68/11/2459/1677293/0362-028x-68_11_2459.pdf by guest on 01 October 2021 22. Laf®, G., S. Cinotti, E. Filimberti, G. Ciabattoni, R. Caporale, F. bacillus endocarditis: report of a case and review of the literature. Marra, L. Melani, A. Grossi, V. Carloni, and P. Gentilini. 1996. De- Rev. Infect. Dis. 8:771±776. fective aggregation in cirrhosis is independent of in vivo platelet 31. Tannock, G. W., K. Munro, H. J. M. Harmsen, G. W. Welling, J. Smart, P. K. Gopal. 2000. Analysis of the fecal micro¯ora of human activation. J. Hepatol. 24:436±443. subjects consuming a probiotic product containing Lactobacillus 23. Marteau, P., and S. Salminen. 1997. Demonstration of safety of pro- rhamnosus DR20. Appl. Environ. Microbiol. 66:2578±2588. biotics. Demonstr. Nutr. Funct. Probiotic Food Newsl. 2:3±8. 32. Zhou, J. S., J. Prasad, P. K. Gopal, and H. S. Gill. 2001. Potential 24. Oakey, H. J., D. W. S. Harty, and K. W. Knox. 1995. Enzyme pro- probiotic lactic acid bacteria Lactobacillus rhamnosus (HN001), duction by lactobacilli and the potential link with infective endocar- Lactobacillus acidophilus (HN017) and Bi®dobacterium lactis ditis. J. Appl. Bacteriol. 78:142±148. (HN019) do not degrade mucin in vitro. Int. J. Food Microbiol. 63: 25. Pelletier, C., C. Bouley, P. Bourlioux, and C. Carbon. 1996. Safety 81±90.