microbiomes, Uncategorized

Akkermansia and Desulfovibrio

This post started out as a PubMed search for papers that mentioned both Desulfovibrio and Akkermansia in the colon. Do they compete with each other? Is one good and the other bad? This post is a followup on the Akkermansia post asking if sulfate or something else is a terminal electron accectpor. Desulfovibrio, as far as I can read so far, uses sulfate as a terminal electron acceptor in its anaerobic respiration. This means hydrogen sulfide production.

They both change with fasting [1]

Periods of fasting and refeeding may reduce cardiometabolic risk elevated by Western diet.
Here we show in the sub study of NCT02099968, investigating the clinical parameters, the
immunome and gut microbiome exploratory endpoints, that in hypertensive metabolic syndrome patients, a 5-day fast followed by a modified Dietary Approach to Stop Hypertension diet reduces systolic blood pressure, need for antihypertensive medications, body-mass index at three months post intervention compared to a modified Dietary Approach to Stop Hypertension diet alone.

two calorie-restricted vegan days (max 1200 kcal/day), followed by 5-days with a daily nutritional energy intake of 300–350 kcal/day,derived from vegetable juices and vegetable broth.


The participants were male and female patients with metabolic syndrome (MetS) according to National Cholesterol Education Program Adult Treatment Panel III (NCEP ATP III)

  1. waist circumference (>94 cm in men and >80 cm in women)
  2. hyper triglyceridemia (>150 mg/dl (1.7 mmol/l) or lipid-lowering medication)
  3. low levels of high-density lipoprotein cholesterol (HDL-C; < 40 mg/dl (1 mmol/l) in
    men and <50 mg/dl (1.3 mmol/l) in women or use of HDL-increasing medication
    (niacin or fibrate)
  4. elevated blood pressure (≥130/85 mm Hg or use of antihypertensive medication)
  5. elevated fasting plasma glucose (≥110 mg/dl or treatment for diabetes mellitus).

Diversity indexes, so what?

Plots 1b and 1c appear to be modified box and whisker plots of sort showing mean, standard deviation, and highest and lowest values. The thickness is some indication of the numbers of participants with that particular diversity score. The text for plot 1d is a direct quote from the publication [1] that contains ambiguous information that the reviewers seem to have missed. Judging from the materials and methods, diversity seems to be defined based on a human gut microbiome gene library containing close to 10,000,000 genes. A second paper referenced in this study described the use of marker gene (MG)-based operational taxonomic units (

  • The Shannon Diversity Index takes into account the relative proportion of the species in a community in addition to number of different different members.
  • Bray-Curtis measures the difference in the populations of two different sites, in this case between two time points of the same individual, across observation times V1–V3. For species found at one time point, but not the other, this species drops out of the 2Cij term. The lower population cont of the marker is used instead of the higher one. In this equation. If the counts are exactly the same, the second term becomes 1 anc BCij becomes 0..
  • Fasting significantly shifts the gut dimensions shown. Axes show Bray–Curtis dissimilarities of rarefied mOTUv2 (marker gene operational taxomic units) OTUs between samples; each participant in the fasting arm is shown as two lines, one red (fasting change), one blue (refeeding change) connected (centered) at the origin for ease of visualization. Axes show fasting and refeeding deltas after one-week intervention and 3-month refeeding. Pseudonym participant ID numbers are shown on the point markers. Transparent circle markers show arithmetic mean position of fasting and recovery deltas, respectively. PERMANOVA test P-values reveal significant dissimilarity (P < 0.05) between samples from each visit V1–V3 in the original distance space, stratifying by donor.

Changes in transcripts for enzymes

Sometimes the technical details of analysis of this publication seem to be a bit much. Should we care more about what Akkermansia and/or Desulovibrio are doing with sulfate in the mucin when we fast? Cysteine biosynthesis transcripts decrease with fasting then increase with refeeding. Is this suggestive of assimilatory sulfate reduction being turned off and then off again? Would the converse of disassemblatory sulfate reduction being turned on during fasting?

Note that this portion of Figure 1g has been edited to only show gene cassettes related to glycolysis, TCA, electron transport chain, sulfate reduction (ASR and DSR) and propionate. d\Red arrows represent fasting effects (V2–V1 comparison), blue arrows refeeding effects (V3–V2 comparison). Bold arrows are significant (nested model comparison of a linear model for rarefied abundance of each taxon, comparing a model incorporating patient ID, age, sex and all dosages of relevant medications) to a model additionally incorporating time point, requiring likelihood test Benjamini-Hochberg corrected FDR < 0.1 and additionally pairwise post-hoc two-sided MWU test P < 0.05. g Gut microbial gene functional modules (KEGG and GMM models analyzed together) significantly enriched/depleted upon fasting/refeeding.

Microbiomes of BP responders

Fasting strongly elevated the abundance of Desulfovibrionaceae and enriched these propionate production modules, indicating that responders suffer a treatable deficit.

  1. Desulfovibrionaceae
  2. Hydrogenoanaerobacterium,
  3. Akkermansia,
  4. Ruminococcaceae GCA-900066225
  5. Hydrogenoanaerobacterium sp.

My take

We have two fasting protocols going on here. The host and the bacteria. We don’t know if there is any causal relationship between the host drop in blood pressure and the microbiome change. Could the same result have been by feeding the host (person) any easy to digest of globular proteins, e.g. milk, and simple sugars with almost no fiber? I find the meabolic changes in the microbiome far more exciting than the phylogenic classification. I find the shift to sulfate metabolism potentially troublesome. What are the consequences of sulfide production, if it occurs? This publication [1] covered a lot of the immune response “ome” that was not included in this post.

The Akkermania/Desulfovibrio Constipation Study [2]

 This study came out of China with the caveat that constipation in an Asian population consuming what is presumed to be a largely Asian diet could present with a different intestinal microbiome than a mixed population in places like the United States consuming a largely Western diet.  Mice striped of their intestinal microbiome with antibiotics were given fecal microbial transplants from healthy human donors without constipation and human donors with constipation.  This post was prepared by doing a key word search of the PDF file of the publication.

Some basic defecation parameters were altered that appeared to be greater on day 15 compared to day 7.  By day 15 the defecation frequency is roughly halved as well as the fecal dry and wet weights.  Note that the units are mg/2hr indicating that the feces simply are not getting larger such there are fewer per hour.  In figure 2 not shown in this post, gastrointestinal transit time increased from about 70 minutes for the healthy human donor feces to over 90 minutes.   Note that the healthy human donors had 6-7 bowel movements a week whereas the constipated donors had 0-2 bowel movements a week. 

Figure 3 compared the 5-HT/serotonin reuptake transporter SERT from the intestines of the mice as well as the colonic cell line Caco2 exposed to fecal filtrates from constipated and healthy feces.  Transcripts for this transporter were increased with exposure to materials in “constipated” feces.   The actual amounts of 5-HT were not all of that different, Figure 4.

There were five human feces donors in each group. Each box is assumed to represent one of them. Desulfovibrio can not be seen in these charts.

Figure 6 examines a less than 50% decrease in Muc-2 immuno staining and PAS staining that reacts with aldehyde groups on the carbohydrate groups of the Muc-2 glycoprotein.  

Fecal microbiota from the mice which received FMT from patients with constipation had a decrease in  Clostridium, Lactobacillus, Desulfovibrio, and Methylobacterium and an increase in Bacteroides and Akkermansia, that these authors claimed to result in the decrease in intestinal barrier.

A previous Akkermansia post has detailed the shedding of assimilatory sulfate reductase genes when A muciphila transfers from strict aerobic conditions to small amounts of O2 found in the mucus interface.  Do these values differ between mice and humans? 

My take

The experimental design was clever enough. The limited reproducibility of the constipated phenotype is interesting. The one big problem with this study is the doubling time of the bacteria. Many are taught that the E coli doubling time is 20 minutes. Sure this is in a nutrient medium. This Harvard site gives doubling time of two strains of E coli in some not so rich media. Here the doubling time is 27 minutes to three hours. Even if we accept the 3 hour doubling time, the time it takes a stool to pass through the colon of a mouse is less than the doubling time. A bacterium may undergo 6-8 to 20 some doublings before the stool is defecated.

Akkermania/Desulfovibrio mucin binding [3]

Desulfovibrio and Akkermansia binding to mucin was examined as it relates to ulcerative colitis.  These authors provided the reader a graphical materials and methods that spurs thought regarding a variety of conditions in which Akkermansia.

Figure 2, not shown, the authors compared different lectins for binding to the carbohydrate fractions of mucins isolated from UC patients.  

Analysis of biopsies collected from each individual indicated that UC mucin had amedian percentage sulfomucin of 39.51% (IQR 32.68%) and controls a median of 57.69%   (IQR 16l74%)  The Inter Quartile Range is the spread of the middle 50% of the sample.  The decreased sulfation of UC mucin was speculated to result from the activities of intestinal bactria.

Three isolates of both A. muciniphila and Desulfovibrio spp. were successfully cultured from different individuals. One isolate of A. muciniphila and Desulfovibrio spp. was isolated from a control patient, and two isolates of both species were isolated from three individuals with UC.

Reference vs UC strains

Table 1 addresses the hypothesis that UC associated strains of A muciniphila and Desulfovibrio chave a higher affinity for mucin.

Healthy vs control mucins

The conclusion is that all isolates of A muciniphila prefer UC mucin to control mucin.  Some Desulfovibrio isolates had higher affinity for UC mucin.  The authors speculated that desulfintion exposes cryptic binding sites on the mucin.

My take

It would seem that Akkermansia would be a way better colonizer than Desulfovibrio of the mucus. It binds with higher affinity according to this clever, but simple assay. We’ve got to remind ourselves that the mucus layer is not strictly anaerobic. We’ve also got to remind ourselves that what is in the feces is not necessarily representative of what is in the mucus layer.

Akkermansia good, Desulfovibrio bad mission aborted

This post started out with the goal of finding peer reviewed literature saying that these two organisms compete for sulfate in our GI lining: one being protective and the other producing a poisonous gas

  1. The Maifeld study demonstrated some very sulfate related changes in the mRNA transcripts with fasting. Actual changes in the sulfur metabalome was not examined… This would have been over the top in an extensive and well performed study.
  2. An increase in Akkermansia seemed to preserve the constipated phenotype, while Desulfovibrio was increased with a faster transit time.. [2] The possibility remains that the faster transit time was the result of an irritated lining…. which was not case according to PAS and mucin immunohistochemistry. [2] That Desulfovibrio is bad and Akkermansia is good is not holding up.
  3. Akkermansia seems to bind mucin with higher affinity than Desulfovibrio…Perhaps we should be less concerned with what is in the feces.


  1. Maifeld A, Bartolomaeus H, Löber U, Avery EG, Steckhan N, Markó L, Wilck N, Hamad I, Šušnjar U, Mähler A, Hohmann C, Chen CY, Cramer H, Dobos G, Lesker TR, Strowig T, Dechend R, Bzdok D, Kleinewietfeld M, Michalsen A, Müller DN, Forslund SK. Fasting alters the gut microbiome reducing blood pressure and body weight in metabolic syndrome patients. Nat Commun. 2021 Mar 30;12(1):1970. PMC free article
  2. Cao H, Liu X, An Y, Zhou G, Liu Y, Xu M, Dong W, Wang S, Yan F, Jiang K, Wang B. Dysbiosis contributes to chronic constipation development via regulation of serotonin transporter in the intestine. Sci Rep. 2017 Sep 4;7(1):10322.PMC free article
  3. Earley H, Lennon G, Balfe A, Kilcoyne M, Clyne M, Joshi L, Carrington S, Martin ST, Coffey JC, Winter DC, O’Connell PR. A Preliminary Study Examining the Binding Capacity of Akkermansia muciniphila and Desulfovibrio spp., to Colonic Mucin in Health and Ulcerative Colitis. PLoS One. 2015 Oct 22;10(10):e0135280.PMC free article
  4. Kang DW, Adams JB, Vargason T, Santiago M, Hahn J, Krajmalnik-Brown R. Distinct Fecal and Plasma Metabolites in Children with Autism Spectrum Disorders and Their Modulation after Microbiota Transfer Therapy. mSphere. 2020 Oct 21;5(5):e00314-20. PMC free article

Leave a Reply