Johne's Disease: Current Trends in Research, Diagnosis and Management

Johne's disease (paratuberculosis) afflicts cattle worldwide and causes significant economic losses. It is also prevalent in goats and sheep and has been implicated in Johne's disease in humans. The book is divided into six sections covering all aspects of the prevalence, management, diagnosis, control and research on Johne's disease.

]ohne's disease is an international animal health problem, and is of particular importance in the southern States of Australia, especially in the major dairy areas of Victoria where approximately 10% of dairy herds are infected.

• Language: English
• Publisher: CSIRO PUBLISHING
• Release date: Jan 1, 1989
• ISBN: 9780643105768

Book preview

Preface

A.K. Lascelles

Impressive efforts by veterinarians with responsibility for animal disease control, using appropriate technologies and epidemiological knowledge developed by research workers in various Institutes in Australia and overseas, have resulted in substantial reductions in livestock production losses from disease. The major clostridial diseases can be controlled by vaccination, and tuberculosis and brucellosis have been virtually eradicated from the national dairy herd. Better mastitis control has been achieved by vigorous application of appropriate husbandry and hygiene measures together with effective use of antibiotics. The Dairy Research Council is optimistic that an effective staphylococcal mastitis vaccine which is in the final stages of field trialling will soon be available to further improve control of mastitis. As these major diseases are eliminated or brought under a reasonable degree of control, diseases which in the past were regarded as less important, assume a high priority. Johne’s disease is a good example.

Johne’s disease is an international animal health problem, and is of particular importance in the southern States of Australia, especially in the major dairy areas of Victoria where approximately 10% of dairy herds are infected.

It is said that the disease is becoming more prevalent. Apart from direct losses caused by the disease, producers can suffer a substantial loss of revenue because health control regulations effectively prevent sale of animals interstate, overseas or to artificial breeding centres from herds even suspected of being infected. The disease also imposes a severe burden on diagnostic laboratories equipped with the laborious, inadequate procedures currently available for establishing a diagnosis.

Opportunities now exist for the development and application of new technologies to vastly improve the precision and speed of diagnosis. This is the major initial aim of research being funded by Dairy Research Council, which already has given Johne’s disease a significant priority in its five-year Research and Development Plan. Council will be prepared to upgrade this priority if it is convinced that the disease is of sufficient importance and that additional research opportunities exist.

I do not believe it unduly optimistic to expect the Conference to achieve the following goals:

-Provide an up-to-date assessment of the incidence and economic importance of Johne’s disease in Australia

-Encourage closer co-operation between officers concerned with diagnosis and control of Johne’s disease in the various State departmental laboratories

-Identify and prioritise research opportunities which show promise of improving diagnosis and control of the disease

-Achieve a Conference consensus which will encourage individual investigators working in different locations to form part of a loosely co-ordinated research effort.

Dairy Research Council, 1601 Malvern Road, Glen Iris, Victoria. 3146. Australia.

The Conference, in terms of inspiration and financial backing, is the joint effort of Dairy Research Council, Victorian Department of Agriculture and Rural Affairs, and CSIRO. In terms of organisation and detailed programming, all the credit goes to Drs. Andrew Milner and Paul Wood. I extend to them, on behalf of the Council and I would believe also all associated with the Conference, our gratitude and congratulations.

Review of recent research studies in the United States related to Johne’s disease with emphasis on diagnosis and control of the disease

R.L. Jones

Introduction

Johne’s disease (paratuberculosis) is a chronic debilitating infectious disease of ruminants that remains virtually undetectable clinically until the onset of a copious, nontreatable diarrhoea and/or chronic weight loss and generalized unthriftiness. Johne’s disease has been recognized in the United States and throughout the world since the early 1900s. However, there has been an increased awareness recently within the livestock industry as evidenced by the interest shown by livestock organizations and major cattle magazines. As a result, the economic impact of the disease, the methods and implications of diagnosis, and the implementation of control and eradication programmes are emerging as major concerns of international proportions.

The purpose of this report is to discuss recent advances in research which are leading to improved identification of Mycobacterium paratuberculosis and offer promise for the development of rapid and accurate diagnostic methods for use in the control of infection. Comprehensive articles (Riemann and Abbas 1983; Chiodini et al 1984) have been published that contain detailed reviews of the literature. Therefore, this article will focus on a few specific areas of recent research in the United States.

Diagnosis

Cultivation of M. paratuberculosis

Mycobacteria isolated from the tissues of animals with Johne’s disease are slow-growing, mycobactin-dependent organisms. It has generally been accepted that mycobactin-dependence qualifies an organism to be identified as M. paratuberculosis, the causative agent of Johne’s disease (Merkal and Thurston 1966). However, mycobactin-dependence can no longer be accepted as the sole criterion for identifying an isolate because many strains of mycobactin-dependent M. avium have been isolated from domestic and wild animals (Matthew et al 1917; Thorel and Desmettre 1982). Biochemical testing of M. paratuberculosis is of limited value since there is variability in reactions between strains (Merkal and Thurston 1966; Chiodini 1986). Likewise, Chiodini and Van Kruiningen (1985) demonstrated considerable variability in the gas-liquid chromatogram of mycobactin-dependent mycobacterial cell wall lipids; a species-specific fatty acid (or pattern) was not demonstrated in their studies. In contrast, Damato et al (1987) reported a distinctive fatty acid peak (not identified) for all M. paratuberculosis isolates that they examined. Serotyping schemes have been ineffective because the rough morphology of most strains causes autoagglutination. Antigenic analysis of M. paratuberculosis indicates that it is closely related to the M. avium (MAC) complex (Collins et al 1983; Mclntyre and Stanford 1986). Hurley et al (1988) have recently compared the DNA-DNA hybridization of M. paratuberculosis with M. avium-M. intracellulare-M. scrofulaceum (MAIS) complex organisms and found that M. paratuberculosis is indistinguishable from some strains of M. avium and M. intracellular. Similar results were obtained by Yoshimura and Graham (1988).

Diagnostic Laboratories, College of Veterinary Medicine and Biomedical Sciences, Colorado State University, Fort Collins, Colorado. 80523. United States of America.

Primary isolation procedures for the recovery of M. paratuberculosis from faeces have been evaluated by several investigators. Early reports (Merkal 1984) that mycobactin J enhanced growth of isolates better than mycobactin P have been difficult to reproduce. In independent evaluations, isolates of M. paratuberculosis have grown equally well in the presence mycobactins J and P (D. Callihan, pers. comm.). Several groups have found that the use of centrifugation in the sedimentation step in culture procedures increases the rate of detection of M. paratuberculosis in faecal cultures (C. Turcotte, S. Bech-Nielsen, and R. Whitlock, pers. comm). Radiometric detection of growth appears to be a promising technique for more rapid identification of M. paratuberculosis in positive cultures (development of this technique is discussed in detail by McDonald et al, p146ff in these proceedings).

Serological tests

The serological diagnosis of Johne’s disease has long been hampered by a lack of specific and sensitive tests. Although a large battery of serological assays have been evaluated, circulating antibodies have generally been considered to be too low in titre, or to difficult to detect, or lack the necessary specificity to be of diagnostic value. The practising veterinarian needs a rapid, economical, reliable test to confirm a case of clinical Johne’s disease. Sherman et al (1984) evaluated an agar gel immunodiffusion (AGID) test as an aid in differential diagnosis of Johne’s disease. In a small study, the results suggested that AGID might be a highly sensitive and specific test in cattle with clinical signs suggestive of Johne’s disease. A similar test is now being commercially produced and marketed for use in practice laboratories. False positive results have not been obtained with the commercial test, but a significant number of clinically affected, culture-positive cows are negative, similar to complement fixation test results (D. Callihan, pers. comm.).

Numerous modifications in enzyme-linked immunosorbent assay (ELISA) methods have been evaluated. Improvements include use of a conjugate specific for IgG1 antibodies (Yokomizo et al 1983), preabsorbing the test sera with a suspension of M. phlei (Merkal 1984), and utilizing an affinity purified peptide antigen (Abbas et al 1983). The activity of various components of these antigens was evaluated by immunoblotting with serum from cattle (Bech-Nielsen et al 1985). Efforts to identify nonprotein antigens are also being pursued. A glycopeptidolipid antigen (to be described below) was identified by our group (Camphausen et al 1985a) and a lipoarabinomannan (LAM) antigen has been described by Sugden et al (1987). The LAM antigen was found to be active in the complement fixation and ELISA tests. This antigen is active in the ELISA at a much lower concentration than other antigens, thus permitting conservation of antigen. Its value as a sensitive diagnostic screening antigen awaits further evaluation; however, it suffers from being a cross reactive antigen with other strains of mycobacteria.

Historically, efforts to identify immunologically active and specific diagnostic antigens have been focused on protein and polysaccharide antigens. To date, truly species specific protein antigens have not been identified within any member species of the Mycobacterium genus although some individual proteins may contain epitopes specific for an individual species (Gillis et al 1985). On the other hand, Brennan (1984) and colleagues have described three immunogenic groups of glycolipid antigens within individual mycobacteria which are relatively simple in structure and provide distinct species and subspecies specificity. The first group, the polar C-mycoside glycopeptidolipid (GPL) antigens, are composed of a monoglycosylated fatty acylated peptide core which is further modified by small variable oligosaccharides. These small oligosaccharide groups (3–5 sugar units) are responsible for the antigenic specificity of all serotypes in the MAIS serocomplex and some rapidly growing mycobacteria including M. chelonei and M. peregrinum (Brennan 1981; Tsang et al 1984). The second group, the lipooligosaccharides, are best represented in M. kansasii. They present the unique immunochemical feature of a nonreducing trehalose substituent at the putative reducing end of the oligosaccharide moiety. The distal end is antigenic. The third group, the phenolic phthiocerol-containing glycosides, were identified in the course of a deliberate search in M. leprae for a species-specific seroreactive glycolipid antigen. The phenolic glycolipid I from M. leprae is highly active in ELISA (Cho et al 1983; 1984).

Schaefer (1965) observed that most nontuberculous mycobacteria have highly immunogenic species- or type-specific antigens and was able to devise a seroagglutination assay for the purposes of identification and classification. His procedure allowed recognition of at least 31 distinct serotypes within the MAIS complex. Serotyping these isolates has been important for epidemiologic and taxonomic purposes. Subsequently, it has been proven that the Schaefer serotyping antigens were the first group of glycolipids mentioned above, the polar C-mycoside GPLs (Brennan 1981). Serotyping of strains has now been supplemented with chemical analysis of cell products. Thin layer chromatography (TLC) of lipid extracts from cells results in unique chromatographic profiles of polar GPLs, distinct for each serotype (Brennan et al 1981). The GPL serotype-specific antigens are also highly amenable to ELISA methodology (Yanagihara et al 1985). The ELISA can be used to assess the antigenicity of glycolipids to detect antibodies in patient serum, and as an adjunct to seroagglutination and TLC, it aids in the identification of nontuberculous mycobacteria.

These advances in mycobacterial glycolipid antigen characterization have provided for identification of serotype- and species-specific antigens. These glycolipids have been purified and structurally characterized. As a result of detailed analysis, the antigenically specific determinants have been identified as residing in small oligosaccharide groups. The precise structure necessary for serological activity has been elucidated using chemically synthesized oligosaccharides and monoclonal antibodies. These discoveries made possible the chemical synthesis of an artificial antigen for the serodiagnosis of leprosy (Cho et al 1984) and identified specific monoclonal antibodies that can detect glycolipid antigen in serum of infected patients (Brennan 1983).

Recently, we commenced a search for serospecific glycolipid antigens within M. paratuberculosis isolates. A major immunoreactive glycopeptidolipid (GPL-I) was isolated and characterized (Camphausen et al 1985a). The glycolipid antigen belongs to the polar C-mycoside GPL family of antigens. GPL-I could readily be detected by TLC; therefore, its presence offered hope as a phenotypic marker for identification of M. paratuberculosis isolates. We later discovered that GPL-I was identical to the GPL which characterized serotype 2 of the MAIS complex (also called M. avium 2) (Camphausen et al 1986). In addition, the TLC pattern of glycolipid antigens from other strains of M. paratuberculosis were analyzed and some were found to contain a major GPL antigen other than GPL-I, and again, a comparison of this GPL with the GPLs from members of the MAIS complex showed that it was identical to the GPL from serotype 8 of the MAIS complex (also called the Davis serotype or M. intracellulare serotype 8) (Camphausen et al 1988). Accordingly, we now propose that some of the strains implicated in bovine Johne’s disease are closely related if not identical to some members of the ubiquitous MAC. However, the principle applies only to some isolates. The majority of mycobactin-dependent faecal isolates do not yield a distinct GPL pattern and accordingly may not be related to the MAIS group or may be rough biovariants that do not produce GPL.

In addition, antibodies specific for GPL-I (serotype 2) and serotype 8 GPL antigens were detected in the serum of some Johne’s disease cows using an ELISA procedure (Camphausen et al 1985b). The presence of specific antibodies in the serum of cattle indicates that GPLs are immunoreactive antigens that cattle recognize by producing specific antibodies. However, cattle appear to produce antibodies that react with the core region of the GPL as well as the specific oligosaccharide determinants (R. L. Jones, unpublished data).

Genetic analysis

The apparent antigenic similarity of laboratory adapted reference strains of M. paratuberculosis with the MAC raised the question of the relationship of these species. Could M. paratuberculosis be a biovariant of the MAC, or are they distinctly different organisms? Therefore, our recent efforts focused on analysis of the genome of these organisms in collaboration with Dr. S. A. Shoemaker. Chromosomal DNA was isolated from multiple strains of M. avium, M. intracellulare, and M. paratuberculosis for restriction fragment length analysis. The DNA’s were cleaved with the following restriction endonucleases: EcoRI, BamHI, SaiI and KpnI. The resulting DNA fragments were electrophoresed through agarose gels, transferred to nylon membranes and probed with a labelled segment of mycobacterial DNA found in multiple copies in the genome. The restriction fragment length patterns were quite similar for most of the M. paratuberculosis clinical isolates. This was in contrast to the patterns seen for the M. avium and M. intracellulare isolates. For these two species, even isolates of the same serotype had different restriction fragment length patterns. The restriction fragment length patterns of the M. paratuberculosis isolates were different from the patterns for the M. avium and M. intracellulare isolates (Shoemaker 1987).

Next, specific DNA sequences which were unique to M. paratuberculosis, M. avium, and M. intracellulare were sought. Such specific DNA sequences could be used as additional markers to define each species, and they could also be used as DNA probes in the development of diagnostic tests. The 16S ribosomal RNA genes were chosen as the target site for identification of sequence differences that would uniquely identify each species (Pace et al 1986). Within these genes there are regions that are highly conserved, while the other regions are more variable in their nucleotide sequences. It is the conserved regions that provide the ability to identify these genes, and the differences in the hypervariable regions that are of particular interest because they are specific for each species. When these hypervariable regions are identified, they can be sequenced for determination of the uniqueness of the ribosomal RNA gene as an identification marker for each organism. Regions considered to be unique in the 16S ribosomal RNA genes of M. tuberculosis H37Rv, M. avium-M. intracellulare, and M. paratuberculosis were sequenced. The 16S ribosomal RNA gene of a recent clinical isolate of M. paratuberculosis and strain 19698 were sequenced and found to be identical to M. avium.

Other research groups have also been studying the genomic structure of M. paratuberculosis. Restriction fragment length analysis has indicated that 31 isolates of M. paratuberculosis, with one exception, produced patterns nearly identical to M. paratuberculosis strain 19698 (Whipple 1987). The pattern produced by mycobactin-independent M. paratuberculosis strain 18 was found to be similar to M. avium serotype 2. Similar results have also been reported by McFadden et al (1987a; 1987b). Using the technique of DNA-DNA hybridization, a high degree of DNA relatedness between M. paratuberculosis and M. avium has been observed (McFadden et al 1987c; Hurley et al 1988; Yoshimura and Graham 1988; Saxegaard and Baess 1988). This data suggests that these organisms are indistinguishable species and should be considered biovariants of the same genomic species. This conclusion is supported by the 16S ribosomal RNA gene sequence data. In summaiy, it appears that M. paratuberculosis is more closely related to M. avium than the relationship of M. avium and M. intracellular. However, restriction fragment length analysis can be used to distinguish these closely related organisms.

Genetic probes

The development of genetic (nucleic acid) probes for detecting M. paratuberculosis in tissue specimens or in faecal specimens could provide a useful procedure for the rapid identification of infected cattle. We propose that DNA probes may be produced that are complementary to the hypervariable regions in the 16S ribosomal RNA genes of a particular species and therefore are specific for that species. There are several advantages of such probes over randomly selected and cloned DNA fragments. Generally, there are several copies of the genes present in the genome. The abundance of ribosomal RNA in the bacterial cells also serves as a hybridization target for the probe. Therefore, the sensitivity of the probe is greatly enhanced. Despite the variability of the target site between different species, it is consistent within a particular species and is unlikely to vary from one isolate to another. Also, the structure and function of the target site and the probe are understood and well defined.

In preliminary clinical evaluations of a genetic probe for M. paratuberculosis, the completion time in the laboratory for diagnosis was reduced to 72 h and 34.4% more of the faecal specimens were positive than by culture (S. Hurley, pers. comm.). False negative test results were not considered to be a problem with the probe. The sensitivity and specificity of diagnosis using the genetic probe have not been determined in field trials. However, in the laboratory, the probe also hybridizes to the DNA of some isolates of M. avium and M. intracellular.

Two of the more serious theoretical limitations of genetic probe diagnosis are the possible lack of sensitivity to detect low numbers of organisms diluted in faeces and the use of radioactive labels on the probe to achieve adequate sensitivity. Recently, a gene amplification technique known as polymerase chain reaction (PCR) has been described which can be used to simplify and speed up the use of genetic probes for diagnostic tests (Marx 1988). The sensitivity of probe assays will be greatly increased by selectively amplifying the desired nucleotide sequence a million-fold. The amount of DNA produced by the PCR is sufficiently large so that radioactive probes will no longer be required for detection. Other less costly detection systems can be employed in laboratories that are not prepared to handle radioactive materials.

Problems in diagnosis of subclinical infection

The major impediment to development and implementation of control programmes is the inability to detect the subclinically infected animal. Some of these infected animals may continue to be negative, even when the ideal antibody detection assay or organism detection assay has been developed. This could be due to the nature of this disease process that results in intermittent shedding of the organism or failure of the infected host to develop an antibody response in the early stages of infection. Also, when highly sensitive assays are developed, one of the trade-offs frequently is a higher rate of false positive results. These false positive tests could be caused by cross-reactions. If highly sensitive direct detection assays are