Monitored therapy of sporadic mycobacteriosis caused by Mycobacterium genavense in Atlantic canaries (Serinus canaria) and Bengalese finch (Lonchura striata)

Mycobacteriosis caused by Mycobacterium genavense was diagnosed for the first time in people with HIV (1). Shortly afterwards, it became apparent that this bacterium is the most common cause of mycobacteriosis in pet birds (11, 12, 27) and those kept in zoological gardens (24, 26), although it may also cause disease in other animal species (14, 21, 34). The bacteria are characterised by very slow growth on selected special media or no growth at all. Therefore, M. genavense has been described as a fastidious organism (6), and is a difficult one to culture; PCR tests are the preferred method of diagnosis of this pathogen (5, 17, 30, 31). Since the 1990s, when this bacterium was first detected, there have been numerous reports of its occurrence in Passeriformes, which show that M. genavense is a serious problem in those birds kept in captivity (23, 27, 30). Studies conducted by Schmitz et al. (30) in flocks of pet birds using the qPCR method showed the prevalence of M. genavense in post-mortem studies as ranging from 3% to 91%, depending on the flock. The faecal-oral route is suspected as the route of infection with M. genavense (19), but the inhalation route cannot be excluded (22, 23). An important role in this regard is played by factors promoting the disease, such as immunosuppression caused, inter alia, by viral infections (22, 30, 31). Mycobacteriosis with an M. genavense agent is treated in humans, although attempts to treat this disease with the same agent have also been made in companion

animals (7). Therapy of mycobacteriosis from M. genavense infection has not yet been described in birds, but attempts at therapy have been made in birds infected with M. avium (16, 18, 28), and the outcomes of these treatments have not been fully satisfactory; neither have they been completely unsatisfactory. The methods of diagnosis and therapy of mycobacteriosis in humans and animals are constantly being improved. Our study was designed to determine the efficacy of the proposed drug combination in the treatment of mycobacteriosis caused by M. genavense, and to assess the suitability of qPCR methods to monitor this therapy.

Studied animals. The therapy was implemented in six male birds originating from a place where mycobacterium infection was diagnosed post-mortem in at least two birds. The total breeding stock of Atlantic canary (Serinus canaria) and Bengalese finch (Lonchura striata) during this period was 45 individuals and they were owned by the same breeder. The birds were kept in cages in the breeder’s apartment. During the breeding season, adult birds were kept in pairs, while outside the breeding season the males were caged individually, and the females in small groups. Young birds were kept in large cages, several per cage. Some of the adult canaries were also in cages on a sheltered balcony.

Birds were selected for treatment based on the results of a microscopic examination of faeces stained using the Ziehl-Neelsen (ZN) method. In the birds selected, acid-fast bacilli (AFB) were found in at least one sample (Table 1). In canaries 3 and 5, a moderate number of Cryptosporidium cysts were also found in ZN faecal stain specimens. The owner of the birds agreed to their hospitalisation, treatment and (in fatal outcomes) subsequent post-mortem examinations. The birds were kept in separate cages in the same room during treatment. Thirteen months after treatment completion, canaries 1, 3 and 4 were placed in cages with one-year-old female red-factor canaries naturally infected with canary circovirus (CaCV) and canary polyomavirus (CaPyV) but originating from another flock. The females were kept with the males for four weeks, or for three weeks in the case of canary 4.

Therapy. The therapy was scheduled to last 24 weeks, except in canary 3, which was treated for 29 weeks. The drug combination was administered twice a day for six days a week and once on the other day. It was given to the birds by 0.8×60 mm feeding tubes into the oesophagus and consisted of clarithromycin (Klacid; Mylan, White Sulphur Springs, WV, USA) at 40 mg/kg b.w., ethambutol (Ethambutol; Teva, Kraków, Poland) at 30 mg/kg b.w. and moxifloxacin (Moloxin; Krka, Novo Mesto, Slovenia) at 10 mg/kg b.w. One day before treatment and weekly thereafter, the body weights of the treated birds were measured, based on which drug dosage was adjusted. Canary 3 was treated again 14 months after completion of the first treatment. The bird received the same drug combination twice a day for seven days a week. The second treatment of this individual was continued until its death on day 87 of treatment.

Humane endpoint. Euthanasia was performed only on birds in a very poor general condition. After prior general anaesthesia with 2% isoflurane (Aerrane; Baxter, Warsaw, Poland), a pentobarbital sodium solution (Morbital; Biowet, Puławy, Poland) at a dose of 1 mL/kg b.w. (10) was administered to the right jugular vein.

Microscopic examination. Microscopic specimens stained using the ZN method were prepared from the faeces of each bird every day for the first two weeks and once a week thereafter.

Blood sampling. Before and after treatment, blood samples in a volume of 0.1 to 0.2 mL were taken from the right jugular vein (29) with a tuberculin syringe with a needle of 0.45×13 mm. From these samples, smears coloured using the Hemacolor method (Merck, Darmstadt, Germany) were prepared and used to establish the approximate leukocyte count (Table 3).

M. genavense quantitative PCR. DNA was extracted from weekly pooled samples of faeces from each bird separately. A GeneMATRIX Stool DNA Purification Kit (EURx, Gdańsk, Poland) was used to extract DNA according to the dedicated protocol. In the second course of treatment conducted in canary 3 only, DNA extraction was performed from faecal samples taken one day before treatment initiation and subsequently 47 times in 87 days of the treatment using a QIAamp DNA Stool Mini Kit (Qiagen, Hilden, Germany), following a two-day protocol (20) including the use of inhibitor-binding tablets (InhibitEX, Qiagen GmbH, Hilden, Germany). Additionally introducing sample homogenisation with glass beads for 10 min on a GeneReady homogeniser (Hangzhou Lifereal Biotechnology Co., Ltd., Hangzhou, China). DNA extraction from the blood and tissues of the studied birds was performed using the GeneMATRIX Tissue & Bacterial DNA Purification Kit (EURx).

DNA extracted from faeces was used in two tests: a SYBR Green qPCR (15) and a TaqMan probe qPCR (30). Both tests were conducted on an Mx3005P real-time PCR device (Stratagene, La Jolla, CA, USA). Reactions were analysed with MxPro software (Agilent, Santa Clara, CA, USA). The primers for both tests and the TaqMan probe were designed based on a M. genavense hypothetical 21 kDa protein gene (5). Samples taken from canary 3 during the second treatment were only tested with SYBR Green qPCR, which employed the forward primer MG 25-s 5′-GAATCCGCTGCTGCTCTG-3′ and the reverse primer MG25-as 5′-TCAATGTAGTCCTGTCCGAAC-3′ (15) with SG qPCR Master Mix (EURx). The qPCR using the TaqMan probe was performed with the forward primer 5′-AAA CAG CGT CAG GAA ATC-3′ and the reverse primer 5′-GTG GGA CGA AGA TGT AGT-3′ as well as the TaqMan probe 5′-FAM-AAC CGC TAT CTA CAT CCG CAG-TAMRA-3′ (30). The PCR assays were performed with Fast Probe qPCR Master Mix (EURx). In both tests, a result was deemed positive at a cycle threshold of under 35, and for the SYBR Green qPCR test, compliance of the melting curve of the tested samples with the standard was also considered indicative of a positive result. The results of the tests for each bird are shown in Fig. 1 A–D and Fig. 2.

Fig. 1

Fig. 2

Cryptosporidium spp. nested PCR. Samples of faeces and organs collected from canaries 3 and 5 were used in a nested PCR for cryptosporidium (3) along with the OptiTaq PCR Master Mix (EURx). The reaction conditions followed those described by Camargo et al. (3). Subsequently, the PCR products were sequenced to determine the parasite species.

PCR tests for Mycobacterium spp. and viruses. The samples of DNA extracted from faeces or organs with the highest qPCR result were selected for each canary. These samples were tested using a PCR for Mycobacterium spp. (36), and the PCR products were subsequently sequenced and identified using NCBI BLAST. The DNA isolates were then compared using Genious Prime software (Biomatters, Auckland, New Zealand). DNA extracted from blood samples and organs (liver, spleen, kidneys, brain, and intestines) was tested in PCRs for circoviruses (9, 32) and polyomaviruses (13), but only organs were tested in an RT PCR for avian bornaviruses (15). The PCR products were sequenced, and species were identified.

Necropsy and histopathological examination. A necropsy was performed as soon as possible after a bird’s death. Fragments of liver, spleen, kidneys, brain, heart and intestines were fixed in 10% formalin and examined histopathologically with routine haematoxylin and eosin and ZN staining.

Clinical observations and survival during and post treatment. Two birds died during treatment. Canary 5 died just two weeks after treatment initiation, while the Bengalese finch died four months after initiation, after a blood collection procedure. One month after treatment completion, canary 2 was euthanised. Canary 1 was euthanised 15 months after treatment completion. In the same year, canary 4 was taken from its cage by a magpie (Pica pica) and unfortunately a necropsy was not performed. Canary 3 died on day 87 of the second course of treatment. Clinical observations and post-mortem changes in the treated birds are presented in Table 2.

In the female kept with canary 3, dwelling on the bottom of the cage and lack of reaction to visual stimuli were observed 8 weeks after their separation. This bird was euthanised and necropsied. Histopathological examination of the organs of this female showed, inter alia, mycobacteriosis of the intestines and spleen. The female kept with canary 1, in which similar clinical signs occurred was also necropsied but no mycobacteriosis was identified. The female kept with canary 4 died two years after their separation; the post-mortem and histopathological examinations showed no mycobacteriosis.

Microscopic examination of faecal samples. Positive microscopy results for mycobacteria were obtained regularly for specimens from canary 1 until the seventh week of treatment, for those from canary 2 until the third week of treatment, and for those from the Bengalese finch in the first week of treatment. In canary 3’s specimens, mycobacteria were found sporadically until the fifth week of treatment; in the second course of treatment, mycobacteria were found irregularly until day 20 of treatment. In specimens from canaries 4 and 5, single AFB were found only prior to treatment. Cryptosporidia were found in all faeces specimens: in canary 5 until death and in canary 3 until the fourth week after treatment initiation.

Microscopic examination of blood. All canaries and the Bengalese finch were diagnosed with heteropaenia, lymphocytosis and monocytosis prior to treatment. In all cases except canary 5, moderate leukocytosis was present before treatment initiation. After or during treatment, leukocyte counts decreased in all birds except canary 5 (Table 3).

Quantitative PCR. For canary 1, 17 positive SG qPCR results and 8 positive TaqMan qPCR results were obtained (Fig. 1A). The SG qPCR yielded four and TaqMan qPCR yielded eight positive results for canary 2 (Fig. 1B). For canary 3, two and four positive results were obtained in SG and TaqMan, respectively (Fig. 1C). Only two positive SG qPCR results were obtained from the tested material of canary 4, in the eighth and fourteenth weeks. No positive results were obtained for faeces in either test on canary 5’s specimens. In the Bengalese finch, seven positive SG qPCR and two positive TaqMan qPCR results were obtained (Fig. 1D). The second test conducted on canary 3’s specimens gave 34 positive results out of 48 tested samples (Fig. 2).

Other PCR tests. PCR products of 1,030 bp in size (34) from samples taken from canaries 1, 2, 3 (before the first and second treatments), 5 and 6 were sequenced. No sequenceable product was obtained from canary 4 in this reaction. All isolates were identified as M. genavense and are presented on a phylogenetic tree (Fig. 3).

Fig. 3

In samples of faeces-isolated DNA from canary 3, the presence of cryptosporidia was revealed up to the 17^th week after treatment initiation; in canary 5, this presence was found in all faeces and intestine samples. Based on the sequencing, the samples were identified as Cryptosporidium galli.

All blood and organ samples revealed the DNA of canary circovirus and polyomavirus. In canaries 1, 2, 3 and all red-factor females, bornavirus was detected in the tested organs (liver, spleen, kidneys, brain, and intestines). Infection with bornavirus was also found in several other canaries from the same breeder who submitted the five treated males, while no bornavirus infection was found in the other tested canaries from the breeder of red-factor females. In the Bengalese finch, the DNA of finch polyomavirus (FPyV) was found in blood and organs.

Canaries and Bengalese finches are good candidates for the proposed treatment. Direct oral administration allows accurate and regular dosing but is feasible only in individual diseased birds housed in small hospital cages (29). An alternative, less stressful method of treatment is to administer drugs with drinking water, but this method does not ensure the correct dosage of the drug (29), especially when it is poorly soluble in water. The authors’ experience shows that canaries and Bengalese finches get used to drug administration through an oesophageal tube as the therapy continues, and immediately after being released from the hand, they return to normal activity.

The basis for the selection of drugs and their dosage were previously conducted studies on the effectiveness of the treatment of mycobacteriosis in budgerigars (Melopsittacus undulatus) (16). The present study demonstrated limited efficacy of therapy of M. genavense infection using the described method. Similar effects were achieved in the previously referenced attempt to treat M. avium in budgerigars (16) and another to treat the infection in Barbary doves (Streptopelia risoria) (28). One of the canaries in the present research was not cured despite 29 weeks of treatment. One year following the treatment, that canary infected a female kept in the same cage for one month. The literature describes a relapse of an apparently cured ferret when mycobacteriosis recurred following a 6-month course of treatment (7). Two birds which shed significant numbers of mycobacteria in their faeces before treatment were cured of mycobacteriosis but developed other fatal diseases. Similar cases have also been described in ferrets treated for mycobacteriosis caused by M. genavense (21). In three birds, we failed to establish the efficacy of therapy because of the death of one before treatment completion and the impossibility of performing a necropsy other. Retreatment of canary 3 did not stop the disease, although its apparent efficacy could be assumed based on the decrease to zero presence of acid-fast bacteria in faeces.

Quantitative PCR tests are a useful tool for monitoring the therapy of M. genavense-induced mycobacteriosis in birds. The assessment of the two PCR methods used allows us to conclude that the TaqMan probe (31) produces more specific results, because the SYBR Green method (17) required result selection based on the melting curve. The TaqMan method was also less time-consuming. However, more positive results were obtained in the qPCR employing SYBR Green. The superiority of the results obtained for samples taken from canary 3 during the second treatment can be explained by the longer sample incubation time as well as the use of inhibitor-binding tablets. Emulsification with glass beads was used in both methods.

Mycobacteriosis in the examined canaries may have been promoted by immunosuppression caused by the concurrent CaCV and CaPyV infections, as it also may have been by FPyV infection in the Bengalese finch. In a retrospective study conducted by Schmitz et al. (31) in birds with mycobacteriosis caused by M. genavense detected post mortem, polyomavirus infection was revealed in 25% and circovirus infection in 35% of cases. However, the birds most susceptible to mycobacteriosis were those infected with both of these viruses. Due to the concurrent CaCV and CaPyV infections in canary 4, it is difficult to determine which virus had a greater effect on the loss of feathers on the tail. Schmitz et al. (31) suggested a significant effect of polyomavirus on moulting in birds. Typical changes in the blood counts associated with mycobacteriosis are leukocytosis, significant heterophilia and monocytosis (25). Leukocytosis was found in the described cases; however, it decreased significantly most probably as a result of the therapy. Monocytosis and moderate leukocytosis, as well as heteropaenia with lymphocytosis instead of heterophilia were also evident in the examined birds. This heteropaenia may have been caused by the co-existing viral infections (35). The treatment used was efficacious against Cryptosporidium galli, but negative PCR results were first obtained only after as long as 17 weeks. Of the drugs in the combination used, clarithromycin was active against Cryptosporidium (8). While mycobacteriosis therapy in birds continues to be controversial (2), studies such as this can help veterinarians in deciding between treatment and euthanasia of infected birds. Euthanasia still tends to be preferred because of the long duration and poor efficacy of therapy, as well as the real risk that mycobacteria pose to human health. Finally, it is worth emphasising that before making a decision on treatment, it is constructive to investigate whether the birds have any coexisting viral infections such as bornavirus which usually develop into fatal disease. The authors are aware that the number of birds used in this study was too small for it to be regarded as rigorous, but even so, it can be concluded that the molecular methods used in the study, which are significantly more sensitive than bacterioscopy and culturing, can be used not only to monitor mycobacteriosis therapy, but also to identify M. genavense and C. galli carriers in pet bird flocks.