In ruminants worldwide Mycobacterium avium subspecies paratuberculosis (MAP) causes Johne's disease 1, which is evocative of inflammatory bowel disease (IBD) in humans 2. As of 2007, 68% of all US cow herds had at least one environmental sample that cultured positive for MAP, rising to 95% in herds of > 500 cows 3. The financial cost of Johne's disease to the agricultural industry, in the USA alone, is estimated to be > $200 million a year.

Humans are continually exposed to viable MAP, as MAP has been cultured from commercially available pasteurized milk in the US 4, and Europe 5 and is found in potable chlorinated municipal water in the US 6 and Europe 7. Although controversial, there is mounting concern 8 and increasingly compelling data that MAP may be zoonotic. 910 Recently, the agents 5-ASA 11, azathioprine 12, 6-MP 1213, methotrexate 13, cyclosporine A, 14 Rapamycin 14 and Tacrolimus 14, that are referred to as "anti-inflammatory" "immuno-modulator" and "immunosuppressants" by physicians, have been shown to cause dose dependent inhibition of MAP in culture. Corroborative evidence that MAP may be zoonotic are that, in humans, the most potent "antiMAP" agents in culture 121314, actually clear MAP DNA from the blood of individuals with Inflammatory Bowel Disease 15. We suggested 111314 that, unknowingly, the medical profession had been treating MAP since 1942, when Nana Svartz introduced sulfasalazine into clinical practice 16.

Introduced in 1967, Monensin 17 is acknowledged as an anti-coccidicidal agent for poultry 181920. Approved as a "Growth Enhancer" antibiotic in the USA, the UK and Europe 21, Monensin accounts for 13% of the total subtherapeutic livestock antibiotic usage in the USA 22. Eukaryotic metabolic effects of Monensin, include inhibition of endosome acidification 23 Na+ ionophore enhancement 24 including action on Leishmania donovani. 25 and possibly prokaryotes 2326. In ruminants the administration of Monensin results in improved energy balance 27, decreases methanogenesis 28, increases plasma urea-N 29, and increased milk production in lactating cows 30. It is of considerable interest that the use of Monensin results in amelioration of pathology 21313233 and results in clinical improvement in animals with Johne's disease 3435.

We hypothesized that Monensin, in addition to its protean anti-coccycidal 181920 and eukaryotic effects 23242536 may additionally inhibit prokaryotes 26, in particular MAP. If correct, ruminant "Growth Enhancement" may in part be consequent to "antiMAP" antibiotic activity in cattle with clinical or sub-clinical Johne's disease. To test this hypothesis, we studied the effect of Monensin on mycobacteria including M. avium and its subspecies paratuberculosis, using our previously validated culture inhibition methods 111314.

This study was conducted as an approved protocol of the Research and Development Committee of the VAMC Bronx approved protocol (0720-06-038.) Inhibition studies were performed on eight mycobacterial strains in culture, as reported. 111314 Five strains were MAP, of which three were isolated from humans with IBD, "Dominic" (ATCC 43545), Ben, (ATCC 43544) 37 & UCF-4 (gift of S. Naser Orlando FL.) Two were MAP bovine isolates, ATCC 19698 (ATCC Rockville MD) & 303 (gift of M. Collins. Madison WI) 38. The M. avium subspecies avium strains (subsequently referred to as M. avium) were ATCC 28291 (ATCC) and M. avium 101 39. To study the effect on the M. tuberculosis complex we used a Biosafety level II surrogate, BCG M. bovis Karlson & Lessel (ATCC 19015.) Agents (all from Sigma, St Louis MO) were dissolved in DMSO 1113 with a final concentration in every Bactec vial, irrespective of the amount of agent in a vial, of 3.2% DMSO.

Quantifying mycobacterial growth and the effect of agents tested, using the radiometric ¹⁴CO₂ Bactec 460 system, has previously been reported in detail 111314. In brief, the daily Growth Index (GI) for each vial is obtained until any vial reaches the instrument maximum of GI of "999." Daily GI's are summated until the day prior to any vial reaching "999." The effect (or lack thereof) of each agent is presented as the percent decrease in cumulative Growth Index (cGI) units (% – ΔcGI) ± SD 11 (when necessary, see individual figures).

For ease of comprehension, data are presented in two ways: For individual mycobacterial strains, data are presented as Figures using the cumulative Growth Index (cGI). For individual chemicals the same data, recalculated as %–ΔcGI 1113, are presented as Tables. Figure 1 = MAP isolated from humans with Crohn's Disease. Figure 2 = Bovine Isolates of MAP from animals with Johne's disease. Figure 3 is M. avium subspecies avium. Figure 4 presents data for BCG. Table 1 is the positive antibiotic control Clofazimine (used in leprosy 40 and clinical trials of Crohn's disease 4142.) The negative controls are the gluterimide antibiotics cycloheximide (Table 2) and phthalimide (Table 3.) The study results on Monensin are presented in Table 4.

In total we performed 29 culture inhibition experiments that included Monensin, of which 21 involved MAP that had been isolated from either ruminants or humans. In every case where Monensin was studied for its effect on MAP, dose dependent inhibition was observed. For brevity and clarity, representative studies are presented. The positive control, clofazimine causes dose dependent inhibition; > 98% – ΔcGI in every strain tested by 1 μg/ml (Figures 1, 2, 3, 4 & Table 1.) A negative control Cycloheximide has no inhibition on any MAP strain at the doses used in these studies (Figures 1 &2 & Table 2.) In contrast, when tested against M. avium ATCC 25291, Cycloheximide has no effect between 1 & 16 μg/ml, but does have 57% – ΔcGI at 64 μg/ml (Figures 3 & Table 2.) Similarly, against BCG, Cycloheximide has no dose dependent inhibition between 1 & 16 μg/ml, but does have 78% – ΔcGI at 64 μg/ml (Figures 4 & Table 2.) Phthalimide, the other negative control, has no dose dependent inhibition against any of the eight mycobacterial strains evaluated (Figures 1, 2, 3, 4 & Table 3.)

Monensin exhibits dose dependent inhibition on MAP in culture, whether isolated from humans with Crohn's disease (Figure 1 & Table 4) or cows with Johne's disease (Figure 2 & Table 4.) Of the human MAP isolates, UCF-4 was most susceptible, 73% – ΔcGI at 1 μg/ml (Figure 1 & Table 4) and Ben least inhibited (51% – ΔcGI at 4 μg/ml: Figure 1 & Table 4.) Of the two bovine MAP isolates, ATCC 19698 (45% – ΔcGI at 4 μg/ml) was less inhibited than 303 (73% – ΔcGI at 4 μg/ml: Figure 2 & Table 4.) Monensin additionally causes dose dependent inhibition on two of the three control mycobacterial strains; ATCC 25291 & BCG (Figures 3 &4 & Table 4.) The single exception to inhibition by Monensin is with the control M. avium subspecies avium strain 101 (Figure 3 & Table 3.)

The "anti-inflammatories" "immune-modulators' and "immune-suppressants" 5-ASA 11, azathioprine 12, 6-MP 1213, methotrexate 13, cyclosporine A, 14 Rapamycin 14 and Tacrolimus 14, all cause dose dependent inhibition of MAP in culture. We suggest that, unlike the majority of antibiotics, which effect only prokaryotes, these agents inhibit both pro and eukaryotes. It is therefore possible that these terms "anti-inflammatories" "immune-modulators' and "immune-suppressants" are actually misnomers. We suggest that these appellations merely report normal secondary eukaryotic physiological effects, consequent to, unknowingly, treating an underlying prokaryotic infection. Specifically we suggest that since 1942, the medical profession has unknowingly been treating MAP infections when using these medications.

Our present data corroborate the culture finding of Brumbaugh et. al. 33. With a single strain of MAP, in one culture experiment, there was no visually detectable growth at 30 days following inoculation. This was interpreted as showing a Monensin Minimum Inhibitory Concentration of 0.39 μg/ml against MAP. In our study, a distillation of 29 different experiments involving Monensin, of which 21 were conducted against five strains of MAP, we show that, in radiometric culture, Monensin causes dose dependent inhibition of MAP, one of two strains of M. avium subspecies avium and a single strain of M. Bovis.

The fact that MAP is the etiological agent in Johne's disease is uncontested. There have been a plethora of observation studies on the "metabolic" effects 26282930 when Monensin is used in cattle. Monensin is called a "Growth Enhancer" by both veterinarians as well as governmental agencies 21. These culture inhibition studies offer a rational explanation for the healing of Johne's lesions 3132 in ruminants with Johne's disease treated with Monensin. As mean prevalence of MAP in US dairy herds steadily increases, currently 68%, 3 it is probable that some of these observations were, unknowingly, made on animals with sub-clinical Johne's disease. We posit that at least some of the "Growth Enhancement" as well as metabolic effects that attends the use of Monensin in ruminants may be the physiological consequence of treating overt or covert MAP infections.

A plausible question is whether the appropriate use of antiMAP agents "cures" Johne's disease? The dose dependent inhibition of Monensin that we demonstrate, are compatible with Monensin being a bacteriostatic, rather than bactericidal antiMAP agent. Our data are therefore compatible with the observations that the use of Monensin ameliorates 32, but does not eradicate or clear MAP from animals with Johne's disease 3133. This would render Monensin analogous to another "antiMAP" agent, 5-ASA where the effect is demonstrably bacteriostatic 11. In our culture system Monensin is not nearly as inhibitory as methotrexate and 6-MP 1213, which clear MAP DNA from the blood of individuals with Inflammatory Bowel Disease 15. We conclude that Monensin cannot be considered as a "cure" for a ruminant MAP infection.

This study does not address the inhibitory mechanism of Monensin on MAP. One possibility is that it is consequent to cell wall destruction, as occurs with coccidia 18. Alternatively Monensin's antiMAP activity may be due to perturbation of obligate intracellular metabolic pathways such a DNA synthesis, as occurs with Methotrexate and 6-MP. Distinguishing between the two modes of action is important. In man, MAP exists in the cell wall deficient form (see 10 for review). In contrast, in ruminants it is the cell wall containing form of MAP that is readily identified 1. We conclude that our present observations do not justify initiating human clinical trials of Monensin in putative zoonotic MAP infections such as Crohn's disease or ulcerative colitis. Such appropriately planned and authorized human trials should only be performed if the antiMAP action of Monensin is unequivocally shown to affect intracellular metabolic pathways of MAP and not its cell wall.

MAP: Mycobacterium avium subspecies paratuberculosis; IBD: inflammatory bowel disease; GI: Growth Index; cGI: cumulative Growth Index; % – ΔcGI: percent decrease in cGI units compared to control growth.

RJG submitted provisional patents based on the hypotheses tested in prior studies. There is no conflict of interest with these Monensin data. STB was a member of the panel of the National Academy of Sciences of the USA that issued the report entitled "The Diagnosis and Control of Johne's disease (ISBN 0-309-08611-6). LS and RHW have no competing interests.

RHW and RJG conceived the experiments. RJG and STB designed the experiments. LS and RJG performed the experiments. RJG, LS, RHW and STB analysed the data. STB and RJG contributed reagents/materials analysis equipment. RJG wrote the manuscript. All authors read and approved the final manuscript.