Effect of frequent consumption of a Lactobacillus casei-containing milk drink in Helicobacter pylori-colonized subjects


Correspondence to: Dr A. Cats, Department of Gastroenterology and Hepatology, The Netherlands Cancer Institute, Plesmanlaan 121, 1066 CX Amsterdam, The Netherlands. E-mail: a.cats@nki.nl


Background : Several studies have reported inhibitory effects of lactic acid bacteria on bacterial pathogens.

Aim : To test whether a drink containing Lactobacillus casei strain Shirota inhibits Helicobacter pylori growth.

Methods : The in vitro growth inhibition of H. pylori was studied when L. casei was added to plates previously inoculated with H. pylori reference strain NCTC 11637. In an intervention study, 14 H. pylori-positive subjects were given Yakult drink (108 colony-forming units/mL L. casei) thrice daily during meals for 3 weeks. Six untreated H. pylori-positive subjects served as controls. H. pylori bacterial loads were determined using the 13C-urea breath test, which was performed before and 3 weeks after the start of L. casei supplementation.

Results : In vitro, L. casei inhibits H. pylori growth. This effect was stronger with L. casei grown in milk solution than in DeMan–Rogosa–Sharpe medium. No growth inhibition was shown with medium inoculated with lactic acid, Escherichia coli strain DH5α or uninoculated medium. Filtration of L. casei culture before incubation with H. pylori completely abolished the inhibitory effect. Urease activity decreased in nine of the 14 (64%) subjects with L. casei supplementation and in two of the six (33%) controls (P = 0.22).

Conclusions : Viable L. casei are required for H. pylori growth inhibition. This does not result from changes in lactic acid concentration. In addition, a slight, but non-significant, trend towards a suppressive effect of L. casei on H. pylori in vivo may exist.


Helicobacter pylori is a micro-aerophilic, spiral, Gram-negative bacterium, which can colonize the hostile acidic environment of the human stomach. In Western countries, the prevalence of colonization with H. pylori varies between 30 and 50%, 1 and in developing countries between 80 and 90%. 2,   3 It is generally accepted that H. pylori can cause peptic ulcer disease, chronic gastritis and gastric cancer in humans. Because of the high prevalence of this bacterium and the strong association with upper gastrointestinal disease, treatment for H. pylori infection has become very common. Multiple drug regimens, mainly consisting of antibiotics and acid suppressive agents, have been tested to achieve the eradication of H. pylori . Despite these combined regimens, eradication is not always effective. New treatment strategies are required to reduce the number of drugs and shorten their administration period, thus improving compliance and decreasing the occurrence of side-effects.

Interference between commensal microbial flora and pathogenic organisms has been described since the end of the 19th century.4 Ever since, probiotics have been used to decrease the colonization of intestinal pathogens in man. Lactobacillus species are probably the best-studied probiotic micro-organisms. Lactobacillus species have the ability to adhere to the stomach wall and to grow under acidic conditions,5 and might be able to reduce the growth of H. pylori in humans. Some in vitro studies have proposed that local pH reduction due to lactic acid production might inhibit H. pylori growth.6–8 Others, however, question this.9

The aim of this study was to investigate whether a readily available commercial preparation containing L. casei inhibits the growth of H. pylori in vitro, and, if so, whether regular consumption of this drink affects the gastric bacterial load of H. pylori-colonized subjects.

Materials and methods

Bacterial strains and growth conditions

L. casei strain Shirota (supplied by Yakult Honsha Co., Ltd, Tokyo, Japan) is a homofermentative, acid-resistant, Gram-positive, rod-shaped, lactic acid-producing bacterial strain, present in the commercially available fermentative milk drink Yakult (Yakult Europe BV, Almere, The Netherlands). L. casei was routinely inoculated from a stock solution (stored at − 80 °C) into 12% skimmed milk solution (Oxoid Ltd., Basingstoke , Hampshire, UK) supplemented with 1% glucose and 1% yeast extract (Oxoid Ltd., Basingstoke , Hampshire, UK). For some experiments, DeMan–Rogosa–Sharpe (MRS) broth (Oxoid Ltd, Oosterzee, The Netherlands) was used as growth medium. Bacteria were grown at 37 °C under micro-aerobic conditions (10% CO 2 , 80% N 2 , 10% H 2 ) for 24 h. H. pylori strain NCTC 11637, obtained directly from National Collection of Type Cultures (NCTC), was routinely inoculated from stock solution (− 80 °C) onto Columbia agar plates supplemented with Dent H. pylori -selective supplement 10 and 7% lysed horse blood (Dent plates), and grown for 48 h at 37 °C under micro-aerobic conditions. Escherichia coli strain DH5α (obtained from Gibco-BRL Life Technologies, Breda, The Netherlands) was grown in Lurea–Bertani medium 11 for 24 h at 37 °C, and served as a negative control.

Determination of growth inhibition

The growth inhibition of H. pylori by L. casei was determined by an inhibition assay, essentially as described by Midolo et al.6 Standard 85-mm Dent plates were dried at 37 °C for 10 min. H. pylori bacteria from a freshly grown confluent plate were suspended in phosphate-buffered saline, pH 7.4, at a density of one McFarland [corresponding to a viable count of ≈ 108 colony-forming units (cfu)/mL], and 200 µL of this solution was then spread equally onto the dried plate. Circular wells (diameter, 8 mm) were then cut out of the agar, with a maximum of six wells per plate. The wells were filled with 80 µL of test solution, and the plate was incubated at 37 °C under micro-aerobic conditions for 72 h, after which the inhibition zone was determined. The inhibition zone was defined as the diameter of the observed growth inhibition (in millimetres) minus the diameter of the well divided by two [(zone diameter − 8 mm)/2]. Unless noted otherwise, the test solution consisted of an overnight culture of L. casei or its supernatant. Supernatant was prepared by centrifuging 1.5 mL of the culture at 16 000 × g for 15 min.

Determination of urease activity

For the determination of urease activity, 10 mL of urea broth solution (Oxoid Ltd., Basingstoke, Hampshire, UK) was inoculated with freshly grown bacteria and incubated for up to 24 h. Urease activity was detected by a pink coloration of the medium.

Subjects and study design

H. pylori -colonized subjects were eligible for inclusion. H. pylori colonization was demonstrated by 13 C-urea breath test. Subjects who had undergone upper gastrointestinal surgery, those with active peptic ulcer disease or concomitant significant neurological, cardiovascular, metabolic, haematological or endocrine disorders, and those who had received proton pump inhibitor or antimicrobial therapy within a period of 1 month prior to the start of the study were excluded.

In a single-centre, open, controlled study, 20 H. pylori-colonized subjects were randomized to receive L. casei supplementation or no additional treatment. For 3 weeks, 14 subjects were provided with 65 mL Yakult drink (milk solution containing 108 cfu/mL L. casei) thrice daily during or after meals. In a control group of six subjects, no dietary changes were made. All subjects were instructed not to use any other commercially available fermented milk drinks 1 week prior to and during the supplementation period. Counting of the returned full and empty bottles was used to test compliance with the protocol.

Before and on the last day of 3 weeks of supplementation, 13C-urea breath test was performed in all subjects. In brief, 100 mg of 13C-urea in a test meal consisting of 150 mL orange juice was ingested during fasting. Breath samples were collected in duplicate in a 10-mL test-tube at baseline and 30 min after ingestion. 13C in exhaled air was measured as the ratio of 13C to 12C using a gas chromatograph-isotope ratio mass spectrometer, and expressed as the difference between 13C/12C after dosing and at baseline (δ13C/12C). The test was considered to be positive if δ13C/12C was above 4.5‰.

The study was approved by the local medical ethical committee. All patients gave written informed consent.

Statistical analysis

Comparison of the inhibition zones of H. pylori growth obtained by L. casei cultured in milk solution and MRS medium was performed using Student's t-test. The number of subjects with a reduction in intragastric urease activity in the L. casei and control groups was compared using Fisher's exact test. Changes in the outcome of the 13C-urea breath test before and 3 weeks after the start of the study were determined in the supplementation and control groups. The Mann–Whitney U-test was used to compare the observed changes between these groups. A P value of < 0.05 was considered to be significant.


Growth inhibition of H. pylori by L. casei (strain Shirota) grown in skimmed milk solution in vitro

Both the viable L. casei bacteria and their culture supernatant clearly inhibited the growth of H. pylori(Table 1). However, the results with the supernatant fraction varied from experiment to experiment, and sometimes we observed bacterial growth at the edge of the wells. These growing colonies were probably L. casei. A Gram stain of the supernatant fraction showed that it still contained L. casei. Therefore, in a subsequent experiment, the supernatant fraction was additionally sterilized by filtration using a 0.2-µm filter (E. Merck Nederland BV, Amsterdam, The Netherlands), and checked for the presence of live L. casei by plating 100 µL of the supernatant on Rogosa plates (E. Merck Nederland BV, Amsterdam, The Netherlands). Whereas the supernatant still contained L. casei (sometimes up to ± 104 cfu/mL), the filtered supernatant did not contain bacteria. Again, we observed a strong inhibition zone for the culture and the supernatant, but not for the supernatant that had been filter sterilized (Table 1). These results suggest that living bacteria are required to inhibit H. pylori. Alternatively, the inhibitory effect may be due to a biocide that is present in limited amounts, and is either unstable or completely bound by the filter used in the sterilization procedure. To discriminate between these two options, an alternative approach to eliminate all L. casei was used. Dent supplement containing antibiotic agents was added to either a fresh L. casei culture or its supernatant, and these were subsequently tested on Dent plates instead of Columbia plates. Dent supplement kills L. casei, even when present in high numbers, but it does not affect the growth of H. pylori. Again, no inhibition of H. pylori was observed (Table 1). Similar results were obtained with sonicated freshly grown L. casei cultures (not shown).

Table 1.  Growth inhibition of Helicobacter pylori by Lactobacillus casei grown in skimmed milk solution, Escherichia coli grown in Lurea–Bertani medium and skimmed milk solution alone
Test solutionIZ (mm) by L. caseiIZ (mm) by E. coli (DH5α)*IZ (mm) by skimmed milk medium
  • IZ, inhibition zone.

  • *

      E. coli grown in Lurea–Bertani medium.

  •   L. casei in 10 mL milk solution grown overnight at 37 °C and diluted 1 : 1 with 10 mL of fresh milk medium (solution 1).

  •  Spinning down solution 1 (1.5 mL) at 20 000  g for 5 min and taking top 1 mL only (solution 2).

  • §

     Filter sterilized solution 2 (0.2 µm filter).

  •  4 µL Dent supplement added to solution 1 and tested on Dent plates.

  • **

     4 µL Dent supplement added to solution 2 and tested on Dent plates.

  • ††

     Solution 2 treated for 1 h with protease K at 37 °C.

  • ‡‡

     Values represent mean ± s.d. of

  • 3 independent assays.

Culture3.9 ± 1.0‡‡0.0 ± 0.00.0 ± 0.0
Supernatant1.7 ± 1.10.0 ± 0.00.0 ± 0.0
Filtered supernatant§0.0 ± 0.00.0 ± 0.00.0 ± 0.0
Culture + antibiotics0.0 ± 0.00.0 ± 0.00.0 ± 0.0
Supernatant + antibiotics**0.0 ± 0.00.0 ± 0.00.0 ± 0.0
Protease K-treated supernatant††7.3 ± 1.37.0 ± 0.06.0 ± 2.0

The results of the experiments described above may still indicate that a relatively unstable biocide is produced by this L. casei strain. In an attempt to characterize the biochemical nature of this putative biocide, the supernatant of a freshly grown L. casei culture was treated with protease K for 1 h at 37 °C. The resultant supernatant inhibited H. pylori growth, but a similar inhibition was also observed when an uninoculated milk solution was treated with protease K and used as a test solution (Table 1). This indicates that the observed inhibition is probably due to protease K activity alone. To exclude the possibility that protease K was responsible for the inhibition, the same solutions were tested again after inactivation of protease K by heating the solutions to 95 °C for 5 min. As a control for the effect of heating, L. casei supernatant not treated with protease K was tested, both with and without heating for 5 min. None of the heat-treated supernatants displayed any inhibition, suggesting that the observed inhibition was indeed due to protease K, and that either living bacteria are required or the putative biocide is heat sensitive.

Growth inhibition of H. pylori by lactic acid

Because none of the treated supernatants displayed growth inhibition of H. pylori, and because these treatments probably do not affect the lactic acid concentration of the supernatant, it is unlikely that inhibition results from the lactic acid production of L. casei. Despite this, we still tested the influence of lactic acid, because of previous claims that lactic acid may be one of the active components causing growth inhibition in the presence of different Lactobacillus strains.6 With increasing concentrations of lactic acid, i.e. 0, 25, 50, 100, 150 and 200 mm, no inhibition of the growth of H. pylori was observed.

Growth inhibition of H. pylori by L. casei (strain Shirota) grown in MRS solution in vitro

It is known that the production of biocines is affected by environmental factors, such as the medium used to grow the lactobacilli. Hence, we repeated some of the experiments using MRS medium instead of skimmed milk solution to grow L. casei. Culture and supernatants from L. casei grown in MRS medium again induced a clear inhibition zone (Table 2). The inhibition zone (mean ± s.d.) was smaller in the culture from MRS medium than in that from milk solution: 2.3 ± 0.3 mm vs. 3.9 ± 1.0 mm (P < 0.01). Thus, with L. casei grown in MRS medium, significant growth inhibition of H. pylori was observed.

Table 2.  Growth inhibition of Helicobacter pylori by Lactobacillus casei in DeMan–Rogosa–Sharpe (MRS) medium, Escherichia coli in Lurea–Bertani medium and MRS medium alone
Test solutionIZ (mm) by L. caseiIZ (mm) by E. coli (DH5α)*IZ (mm) by MRS medium
  • IZ, inhibition zone.

  • E. coli grown in Lurea–Bertani medium.

  • †  L. casei in 10 mL MRS solution grown overnight at 37 °C in 5% CO 2 , followed by the addition of 10 mL of fresh milk medium (solution 1).

  • ‡ 

    Spinning down solution 1 (1 mL) at 14 000  g for 5 min.

  • § 

    Values represent mean ± s.d. of

  • 3 independent assays.

Culture2.3 ± 0.3§0.0 ± 0.00.0 ± 0.0
Supernatant2.3 ± 0.30.0 ± 0.00.0 ± 0.0

Urease activity of L. casei (strain Shirota)

Although most Lactobacillus species are urease negative, some have been reported to produce urease. If L. casei strain Shirota produced urease, it would interfere with the 13C-urea breath test used in the in vivo part of this study. Therefore, we investigated whether urease activity was present in L. casei strain Shirota. Even on prolonged incubation (> 24 h), L. casei strain Shirota did not change the colour of urea broth, indicating that it has no intrinsic urease activity. As a positive control, we used H. pylori strain NCTC 11637, and as negative controls we used both E. coli strain DH5α and a urease-negative isogenic variant of H. pylori NCTC 11637, constructed in our laboratory by allelic exchange, essentially as described by Ferrero et al.12

Effect of L. casei on urease activity in vivo

All 14 H. pylori-positive subjects (six males, eight females) to whom L. casei was given were fully compliant. The age (mean ± s.d.) was 51 ± 13 years (range, 32–73 years). At the start of the study, two subjects expressed no complaints, five had non-ulcer dyspepsia, two had constipation, three had endoscopic signs of gastritis or bulbitis, one had endoscopic signs of gastro-oesophageal reflux and one had a history of duodenal ulcer disease. A group of six H. pylori-positive subjects (two males, four females) with an age (mean ± s.d.) of 44 ± 9 years (range, 34–57 years) served as controls. In this group, three subjects had non-ulcer dyspepsia, one had symptomatic gallstone disease, one had a history of duodenal ulcer disease and one asymptomatic subject had a family history of gastric cancer.

Urease activity, as determined by 13C-urea breath test, was similar in both groups at baseline (Figure 1). The median (range) urease activity was 31‰ (14–84‰) in the L. casei supplementation group and 47‰ (19–104‰) in the control group (P = 0.15). Urease activity decreased in nine of the 14 (64%) subjects with L. casei supplementation and in two of the six (33%) controls (P = 0.22). The median (range) urease activity after L. casei supplementation was 24‰ (13–76‰) vs. 31‰ (22–131‰) in the control group (Figure 1). The complaints expressed prior to the study decreased in five subjects in the L. casei group, whereas one subject complained of diarrhoea during the study. No changes were reported in the control group.

Figure 1.

13 C-Urea breath test in H. pylori -positive subjects before and at the end of a 3-week period with and without Lactobacillus casei supplementation.


Colonization with H. pylori causes a wide range of upper gastrointestinal disorders in humans. Several of these pathophysiological changes can be reversed by the elimination of the bacterium, which can be achieved through the administration of a combination of antimicrobial and acid-reducing agents. Unfortunately, eradication therapy is not always successful and may even induce several side-effects, such as pseudomembranous colitis. This develops through disruption of the ecological equilibrium of the intestinal micro-environment, allowing the emergence and development of the pathogenic Clostridium difficile. Probiotic organisms may stabilize or restore the endogenous intestinal microflora. This may result from an aspecific action against pathogenic organisms, such as competition for nutrients or adhesion sites on epithelial cells, as well as through the specific production and secretion of antibacterial substances. Lactobacillus species have demonstrated a beneficial effect in various conditions associated with a disrupted gastrointestinal micro-environment, such as traveller's diarrhoea,13, 14 acute rotavirus diarrhoea15 and antibiotic-associated diarrhoea.16, 17 During the 1990s, the inhibition of H. pylori by several Lactobacillus strains was reported in vitro and in mice. Studies evaluating the probiotic mechanisms of Lactobacillus on H. pylori demonstrated a reduction of adhesion of H. pylori to (gastric) epithelial cell lines in the presence of Lactobacillus species.18, 19 In addition, other studies suggested that inhibition might be due to the production of lactic acid, its consequent pH reduction, or both.6, 7, 20

Several Lactobacillus species have been shown to have consistent probiotic effects,21 and, of the many available Lactobacillus species, we decided to use L. casei strain Shirota because it is commercially available for consumption as a fermented milk drink in more than 20 countries. Our data show that L. casei inhibits H. pylori growth in vitro. This corresponds with other recent data.9, 22 Similar results have been obtained with other Lactobacillus species, including L. salivarius,8L. acidophilus6, 7, 9 and L. bulgaricus.6 To our knowledge, this is the first time that the requirement for viable L. casei has been demonstrated, indicating that the growth inhibitory effect is due to an (unstable) biocine, rather than to lactic acid or pH. A similar observation of lactic acid- and pH-independent growth inhibition of H. pylori has been made for L. acidophilus.9 The involvement of a proteinaceous compound has also been suggested in a recent study, which reported that the inhibitory action of L. acidophilus on the growth of H. pylori was associated with the spontaneous autolysis of L. acidophilus that occurred after 24 h of culture.20

The growth of L. casei in skimmed milk solution augments the inhibitory effect of L. casei on H. pylori. It has been described that milk enhances the survival of Lactobacillus in gastric juice and increases the adhesion to intestinal cells in vitro.23 This may play a role in our study as well. Alternatively, the bacterium might produce more of the active compound under these conditions.

Studies evaluating the probiotic effects of Lactobacillus strains on H. pylori colonization in humans are sparse. Recently, a placebo-controlled study indicated a clear decrease in H. pylori density, inflammation and activity scores, determined in histological sections, following a 3-week intake of acidified milk containing L. johnsonii in H. pylori-positive subjects.24 In an uncontrolled study by Michetti et al., a decrease in urease activity, assessed by 13C-urea breath test, in 16 of 20 H. pylori-positive subjects treated for 2 weeks with a whey-based culture supernatant of L. acidophilus was reported.18 In a non-blind, cross-over study in 31 Japanese H. pylori-positive subjects, a slight but significant decrease in urease activity was observed shortly after the consumption of L. gasseri.25 However, we also observed a decrease in urease activity in our control group. The reliability of the 13C-urea breath test has been convincingly established, with a sensitivity and specificity in the range 95–100%.26, 27 In addition, the correlation between H. pylori density and severity of inflammation, assessed in histological slides, and the amount of urease activity, determined by the 13C-urea breath test, is well documented.28–30 Thus, there is some inter-individual variation in urease activity over time. In our study, frequent consumption of an L. casei-containing milk drink resulted in slightly decreased H. pylori colonization. However, probably partly due to the fluctuation in urease activity in the rather small control group, this decrease was not statistically significant. As the eradication of H. pylori following the consumption of Lactobacillus has not been reported,18, 25, 31 we believe that the potential role for probiotic Lactobacillus lies in the facilitation of eradication strategies against H. pylori. This is underscored by two recent studies which show the beneficial effects of pro-biotic Lactobacillus strains added to the standard 7-day triple therapy against H. pylori.32, 33 In the first study, 120 H. pylori-positive patients were randomized to receive rabeprazole, clarithromycin and amoxicillin, either with or without a lyophilized and inactivated culture of L. acidophilus thrice daily.32 Although the eradication rate in the control group was rather low (72%), a significantly higher eradication rate was observed in the L. acidophilus-supplemented group (88%). It is not clear why non-viable bacteria achieved this beneficial effect. In the other study, the effect of Lactobacillus strain GG or placebo, added for 14 days to triple therapy with rabeprazole, clarithromycin and tinidazole, on the gastrointestinal side-effects associated with eradication therapy was determined in 60 H. pylori-positive subjects.33 Diarrhoea, nausea and taste disturbance, assessed by questionnaire, were significantly reduced in the Lactobacillus strain GG-supplemented group. These human studies underscore the potential benefit of probiotics, such as Lactobacillus strains, in the treatment of H. pylori-associated gastrointestinal disorders.

In conclusion, viable L. casei, especially when grown in the presence of milk, are capable of inhibiting the in vitro growth of H. pylori. This effect is not explained by lactic acid production, but by close contact of live L. casei cells with H. pylori in the medium. Although the effect of L. casei on H. pylori growth and activity in humans may be promising, additional larger scale studies are needed.


This study was sponsored by Yakult Europe BV, Almere, The Netherlands.