Molecular Detection and Parasite Load Determination of Leishmania donovani in Dogs at Humera and Sheraro, Western Tigray, Ethiopia

The authors have withdrawn their manuscript owing to the article was submitted and made public without the full consent of all the authors. Therefore, the authors do not wish this work to be cited as reference for the project. If you have any questions, please contact the corresponding author.


Author Summary
Leishmaniasis is a neglected vector-borne disease of tropics and subtropics endemic to many countries across the world.It is caused by obligate intracellular protozoa of the genus Leishmania and is a disease of both humans and animals.In the present study areas, Humera and Sheraro, human visceral leishmaniasis is endemic.However, the source of infection is not yet well known.There are some seroprevalence reports of the parasite in some animal species in these areas but not confirmed with parasitological and or molecular techniques.In this study, the authors examined blood samples from dogs living at human visceral leishmaniasis endemic area with serological and molecular (quantitative PCR) techniques.DNA of the parasite was detected in some clinically symptomatic and asymptomatic dogs suggesting dogs could be potential reservoirs for the parasite at the area.The attempt for parasite load quantification was not successful due to low level of DNA detected.Confirmation of the presence of the parasite in dogs will help in designing appropriate prevention, treatment and control measures of the parasite in the study areas.

W I T H D R A W N
s e e m a n u s c r i p t D O I f o r d e t a i l s

Introduction
Leishmaniases are a group of diseases with a broad range of clinical manifestations caused by several species of parasites belonging to the genus Leishmania (Family: Trypanosomatidae).
Leishmania parasite, a haemo-flagellate protozoan organism, is exclusively transmitted by the bite of female sand-fly of the genus Phlebotomus or Lutzomyia (1).It is a vector-borne tropical disease with major public health concern in poor countries and yet not given the attention it deserves.
Leishmania parasite gets into human population when human, sand fly and the reservoir hosts share the same environment (2).Leishmania is transmitted to humans and other mammals by the bite of an infected sand fly vector (3).There are two main sources of human leishmaniases; zoonotic leishmaniases, in which the reservoir hosts are primarily wild animals, commensals or domestic animals, and anthroponotic leishmaniases, in which the reservoir host is primarily human (4).
Leishmania (L.) donovani and L. infantum are species of the L. donovani complex and are the main causative agents of visceral leishmaniasis (VL).L. infantum is responsible for zoonotic VL, with dogs as the main reservoir hosts in the Mediterranean area, the Middle East, Asia, and South America (5).On the other hand, L. donovani is thought to be anthroponotic, and responsible for VL both in the Indian subcontinent and eastern Africa (6).
According to world health organization (WHO) and Pan American Health Organization (7), leishmaniasis can present in three clinical forms.Cutaneous leishmaniasis is one form of the disease, manifested by localized or multiple skin ulcers.Visceral leishmaniasis is the most severe form of the disease manifested by fever, weight loss, inflammation of the spleen and liver, and anemia, and can be fatal in 90% of cases, if not properly treated).Mucosal leishmaniasis is another form of the disease in which lesions occur on mucous membranes of the nose, mouth, throat cavities, and surrounding tissues.

W I T H D R A W N s e e m a n u s c r i p t D O I f o r d e t a i l s
Globally, both cutaneous and visceral leishmaniasis are endemic to many countries (8).An estimated 50,000 to 90,000 new VL cases occur each year across the globe (9).More than 90% of the global reported VL cases occur in six countries: India, Bangladesh, Sudan, South Sudan, Ethiopia and Brazil (8).
Leishmaniasis is endemic to Ethiopia.Plain areas of Humera (western Tigray) and Metema share 60% of the burden (10).According to a recent "review by (11)", the overall random pooled prevalence of leishmaniasis in Leishmania endemic areas of Ethiopia from 34 previous studies done on 52,706 humans and animals was 16% (95% CI: 12%-20%) in humans and 21% (95%, CI: 15%-27%) in animals with major occurrence in the northern part of Ethiopia.The author of this review also stated, VL as the most dominant type of leishmaniasis in Ethiopia (11).
In different areas in Ethiopia, anti-Leishmania donovani antibodies detected in dogs (12), domestic animals (13) and rodents (14).In northern Ethiopia, studies have detected L. donovani antibodies in domestic and wild animals (15).In a study conducted by (13) on exposure to Leishmania species to domestic animals in Humera, Sheraro and Adis alem, 18.9 % of animals were seropositive for anti-L.donovani IgG with highest sero-positivity in dogs (55.9%) followed by sheep (10.5%) and goat (10%).However, whether dogs could potentially be a reservoir for transmission of the parasite to sand flies or not is yet not clearly stated.The hypothesis was, L. donovani can be detected in clinically symptomatic and asymptomatic dogs in an area endemic for human VL caused by L. donovani and dogs could have sufficient parasitemia /skin parasite load to be potential reservoirs for transmission to sand fly.
Hence, detecting the parasite DNA, determining the parasitemia level and skin parasite load in dogs could suggest parasite transmission possibility from clinically diseased and asymptomatic dogs to sand flies.Therefore, the primary objective of the study was to detect the L. donovani parasite in symptomatic and asymptomatic dogs living in human VL endemic areas, Western Tigray through real time quantitative PCR (qPCR) and quantify the parasite load.

W I T H D R A W N
s e e m a n u s c r i p t D O I f o r d e t a i l s

Study Area
The study was conducted in two districts of western Tigray (Humera and Sheraro) (Fig. 1) in the northern part of Ethiopia from October 2019 to April 2020.Humera has a hot semi-arid climate with fertile land used for agricultural production.Human visceral leishmaniasis is endemic to the area (11).Human visceral leishmaniasis incidence is more common during weeding and harvesting season (July-December) than in dry season (January-June) (16).Previous research reports have also shown the detection of anti-L.donovani anti-bodies in various domestic animals among which dogs are the main ones in these areas (13,15,17).Moreover, phlebotomine sand flies are believed to be present especially in Humera where they prefer to live in hot low land areas (18).Sheraro is also near to Humera, in a relatively higher elevation, some 1246 meters above sea level.

Study Design
Cross sectional study was employed in dogs living in human VL endemic areas with the objective of detecting L. donovani and determining the parasitemia and skin parasite load as an indication of reservoir potential for transmission to sand fly.Study sites were selected purposively based on previous disease prevalence and incidence rates.Sampling period was in the autumn (October-November), 2019.

Sampling
A total of 90 (Humera, n= 44; Sheraro, n=45; Mekelle, n=1) dogs were selected from 303 dogs that visited veterinary clinics for rabies vaccination (in Humera), scheduled deworming (in Sheraro) and in Mekelle (for negative control).Thirty six (36) dogs (n=20 from Humera and n=16 from Sheraro) with suspected symptoms ((weight loss, alopecia, dermatitis, abnormally long or brittle nails) (19), (20), ( 21)) were purposively selected.The remaining 53 dogs with no clinical symptoms were selected through systematic sampling frame of one every five dogs.One additional apparently healthy dog selected from Mekelle for negative control.Selected dogs were registered on structured dog information recording format (S2 Appendix).All dogs were owned, local breed and always spent 24 hours outdoor.

W I T H D R A W N
s e e m a n u s c r i p t D O I f o r d e t a i l s Six milliliter (ml) blood sample was withdrawn from cephalic vein of all selected dogs (n=90) with syringe, immediately reconstituted to 10 ml blood collection vacutainer tube with EDTA and processed as per the sample processing layout (S3 Appendix).Additional lesion scraping samples were also collected from 14 dogs using sterile blade, placed in 1.5ml capped tubes with no buffer.

rK39ITLeish Dipstick Test
After separation of plasma from whole blood through centrifugation at 1500 revolution per minute (rpm) for 10 minutes (S3 Appendix), rk39ITLeish dipstick test (from InBios International, Inc. Seattle, WA USA 98104) which is Kalazar Detect Rapid Test used for leishmaniasis diagnosis in humans was done on plasma samples of all 90 dogs to detect antibodies of Leishmania donovani parasite according to manufacturer's test procedure.Some studies have shown that the test has been done in dogs before (22,23).Test results were scored qualitatively according to manufacturer's result interpretation guideline as indicated in (Fig. 2).

Quantitative Real Time PCR (RT-qPCR)
Quantitative real time PCR was done for all DNA extracts and controls.Amplification was made in a final volume of 25l PCR reaction targeting kinetoplast minicircle DNA (kDNA) of Leishmanaia donovani parasite (25).PCR mixture components included in the PCR reaction and sequence of primers are presented in (table 1 and 2

W I T H D R A W N
s e e m a n u s c r i p t D O I f o r d e t a i l s

Clinical Examination, rK39 Serology
Of the 89 clinically examined dogs, 36 were clinically suspected.Clinical symptoms noticed were skin lesions (alopecia, dermatitis), weight loss, emaciation and long and brittle nails.The rest look apparently healthy.Only 6.74% (6/89) were positive for rK39ITLeish dipstick test.The sample from non-endemic area of one dog was negative for rK39ITLeish dipstick test and PCR.

Technical validation of no template controls (NTCs) in extraction and PCRs
PCR signal was detected in two no template control (NTC) wells included in extraction and PCR with ct values of 18.1 and 40.32 respectively.14 NTC wells in extraction and 14 NTC in PCR show no amplification detection (Fig. 7).

Technical validation of negative control and negative extraction controls in PCRs
No signal detected in all the five negative control (2l DNA of whole blood from dog of nonendemic area) and five negative extraction control (extract of distilled water) included in all PCRs (Table 5).

PCR results of dog samples
For real time qPCR analysis, 47 dog samples including seropositive samples were used.PCR signal detected in 3/47 dog samples (3.37%); in whole blood, buffy coat and plasma of n=2 dogs and in whole blood of the other dog.No signal detected in lesion scraping samples.In positive dog samples, there was variation in detection and when detected, there was variation in amount of DNA between sample types (whole blood, buffy coat and plasma) of different PCR runs.
Positive samples were spiked with 10fg DNA in PCR tube to see if the variation was due to PCR inhibition.Cycle of threshold (Ct) values and DNA amount (in fg) of spiked samples were compared with ct value and DNA of 10 fg in standard curve (Table 6).

Sample Positivity and Correlation in Different Diagnostic Techniques
Of the total 89 dogs, 36 were clinically suspected and six positive for rK39ITLeish dipstick test.47 dog samples were employed for RT-qPCR analysis from which PCR signal was detected in only three dogs (two positive for rK39ITLeish dipstick test).PCR analysis was not done in the remaining 42 dogs due to limited budget.Considering RT-qPCR as reference test, Diagnostic accuracy of rK39ITLeish dipstick test was 89.36% (95% CI: 76.90%-96.45%)with sensitivity and specificity 66.67% (95% CI: 9.43%-99.16%)and 90.9% (95%, CI: 78.33%-97.47%)respectively.Summary of positivity and correlation in different diagnostic techniques are summarized in (Tables 7 and 8).

Sample Type and PCR Positivity
Positive PCR signal was detected in three dog samples of whole blood and buffy coat and in two plasma samples with varying DNA amount.There was no signal detected in skin lesion scraping samples (Fig. 8).

Discussion
In the present study, a total of eighty nine (89) dogs from human VL endemic areas and one dog from non-endemic area (as negative control) used for leishmaniasis diagnosis with clinical examination and rK39 serology.RT-qPCR employed on samples of 47 dogs.
Clinically, 36 of the dogs had Leishmania related symptoms.Skin lesions (alopecia, dermatitis) were major clinical symptoms (in 18/36) followed by long and brittle nails (in 11/36).Six dogs had skin lesion, weight loss and long and brittle nails; one dog with skin lesion and emaciation.
However, only one dog from those with Leishmania related clinical symptoms were positive in RT-qPCR.
The seroprevalence of 6.74% detected in this study using rK39ITLeish dipstick test was lower than the 14.8% (from N=162) reported by ( 15) using kalazar detect test (KDRT) but slightly higher than 4.35% (from N=46) observed by (27) using ITLeish rapid detection test.This could be due to differences in sensitivity and specificity of serological tests (28), infection status of the patient (23) and differences in sample size.Another reason could also be due to differences in the type of test kit used where rK39ITLeish dipstick test used in this study was the one designed for leishmaniasis diagnosis in humans.Herrera and colleagues suggested that designing of tests with antigens of the circulating strains in humans and dog would improve diagnostic utility (22).
Presence of rK39 antibodies in dogs in the present study areas indicated possible contact of L.
donovani with dogs.However, this doesn't always reflect true infection due to i) low specificity of the serologic test and or presence of other cross reacting parasites in sampled dogs (29), ii) inability of the serologic test to differentiate active infection from cured as circulating antibodies may persist longer after cure (30).
The most important finding of the study was detection of L. donovani DNA in dog samples (whole blood, buffy coat and plasma) through RT-qPCR.In order to detect and quantify, standard PCR curve was made for each PCR run.Only two of the 36 dogs with clinical symptoms were positive for rK39 antibodies whereas, of the apparently healthy 53 dogs, 4 were positive with detectable rK39 antibody.Absence of clinical symptoms in rK39 antibody positive dogs could either be due to recovery from infection but persistent circulating antibodies or resistance to the parasitic infection.Kalayou and colleagues also reported detection of Leishmania donovani antibodies in 14.8% asymptomatic dogs in the same study area (15).
Of clinically suspected 36 dogs, only one dog tested positive for RT-qPCR.The other two PCR positive dogs were asymptomatic.Reasons for asymptomatic infection could be that; dogs could be resistant to the parasite, diagnosis could have been done at early infection before clinical symptoms develop.Therefore, asymptomatic dogs can still have parasites.Observed clinical symptoms in dogs tested negative for RT-qPCR could be due to other diseases as other infections can cause similar clinical presentations too.This shows that, all symptomatic dogs are not truly infected with the parasite as Leishmania infection does not show pathognomonic clinical symptoms.

W I T H D R A W N
s e e m a n u s c r i p t D O I f o r d e t a i l s Only 33.33% (2/6) dogs with rK39 antibodies tested positive by RT-qPCR.Four seropositive dogs tested negative for PCR.This could be due to the reason that dogs recovered from infection but still have circulating antibodies tested positive for rK39.This is the major limitation of serological tests; inability to differentiate active infection from relapsed ones (30).
In conclusion, Leishmania donovani DNA was detected in one symptomatic and two asymptomatic dogs living in human VL endemic areas.DNA was detected in different sample types; whole blood, buffy coat and plasma with varying DNA concentration but not in lesion scraping.Parasitemia and skin parasite load were not calculated because of high ct values of positive dog samples in repeated PCR runs and sample types.Small sample size of dogs used in this study and use of rapid diagnostic test not designed for dogs may partially contribute for some inconclusive results.Use of enough sample size of dogs, and follow up study on confirmed canine leishmaniasis cases from the area, xenodiagnosis, looking at skin parasite load which may be more relevant for infectivity to sand-fly is suggested in order to explore relevance of dogs as reservoir for transmission of Leishmania donovani in Ethiopia.

W I T H D R A W N
s e e m a n u s c r i p t D O I f o r d e t a i l s

Fig 2 .
Fig 2. Kalazar Detect Rapid test interpretation appearance of test line and control line is interpreted as positive, control line with no visible test line interpreted as negative, no test line and control line indicate invalid test, abbreviations: L=test line, C=control line

Quantitative PCR (RT-qPCR) 3 . 2 . 1 .
Technical validation of standard PCR curvesSamples were analyzed in five PCR runs.Standard curves made from Serial dilution of known Leishmania donovani DNA (BPK282/oclo4/12) starting from 10000 femtogram (fg) down to 0.01 fg in steps of 10 were included in all the five PCR runs.Efficiency of the PCR is 2.04 with limit of detection being 0.01fg DNA in all PCRs.The linear relationship between the logarithmic DNA amount in femtogram and ct values is Y=aX+b, i.e -3.2262X+38.311with an R square value of 0.994.Details of linear dynamic range and connected data points in five PCR runs are presented in (Fig.4 and 5) respectively and used as a reference for quantification in all PCRs.

Fig 4 .Fig 5 .
Fig 4. Linear regression of ct values as a function of DNA amount in femtogram used in all PCRs.It is used as a reference for quantification in all PCRs.Data points within the circle above are detected DNAs that cannot be quantified.

Fig 6 .
Fig 6.Technical validation of positive control in PCRs: Blue diamonds in the above figure represent input DNA before extraction in all extraction rounds, red squares represent output DNA in PCR.

Fig 7 .
Fig 7. PCR results of no template controls.Ct value of 50 is an artificial value for negative resultsBlue diamonds in the above figure represent for negative results of no template controls, red squares represent positive results of no template control

Fig 8 .
Fig 8. Sample type and PCR positivity by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.It is made

Table 2 :
Primers and Probe

Table 3 :
Light cycler 480 computer program rK39ITLeish dipstick test.RT-qPCR results were analyzed using linear regression for calculating efficiency of PCR, R square value, limit of detection, slope, intercept, and to construct the linear relationship between logarithmic DNA and ct values.Kappa test agreement was also used as a measurement of agreement between diagnostic test results.Sensitivity and specificity of diagnostic techniques were calculated using MedCalc's diagnostic test evaluation calculator taking RT-qPCR as a reference.
All data related to diagnosis results were stored in Microsoft access and database was created (S4 Appendix) for analysis.Excel was employed to summarize results of clinical examination and s e e m a n u s c r i p t D O I f o r d e t a i l s

Table 4 :
Summary of input DNA before extraction, output DNA in PCR, and kit used

Table 5 :
PCR results of negative control and negative extraction controls Negative control and negative extraction controls, all tested negative

Table 6 :
DNA amount in fg in 2l of positive dog samples, Ct values, DNA amount (in fg in PCR) before and after spiking by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.It is made All PCR results with less than 1fg DNA amount are described as <1fg.In the table, ct values are written first followed by DNA amount in fg in brackets.Abbreviations: ND=not detected, PCR=polymerase Chain Reaction, fg=femtogram, ct= cycle of threshold s e e m a n u s c r i p t D O I f o r d e t a i l s

Table 7 :
Summary of corelation between different diagnostic technique results s e e m a n u s c r i p t D O I f o r d e t a i l s

Table 8 :
Summary of positivity of PCR positive samples in clinical exam, rK39ITLeish dipstick test (36)35)date the extraction process and possible contamination, positive control, negative control and negative extraction controls were included in each extraction round.All positive controls included in each extraction round become positive in PCR but with reduced DNA yield than input DNA in extraction.With inclusion of 100pg DNA positive control in extraction and assumed 80% extraction efficiency in 50l elution buffer from which 2l DNA template included in PCR, an output of 3.2pg DNA was expected in PCR.However, only 1070fg and 1890 fg DNA was detected in QIAamp blood mini kit and 63.7 fg DNA in QIAamp mini kit.QIAamp mini kit is originally designed for DNA extraction of tissue samples but in this study, it was used for DNA extraction from whole blood of positive control and this could be the reason for the low DNA yield in PCR (63.7 fg DNA).All negative and negative extraction controls tested negative.No template controls (NTCs) included in PCRs tested negative except in two NTCs with possible contamination.These two positive NTCs could be due to either possible contamination in extraction room or error in pipetting from serial dilution of higher concentration.Three of the 47 dog samples tested positive for RT-qPCR with varying DNA concentration in different PCR runs.This is the first report of molecular detection of L. donovani DNA in dogs in the study areas where human VL is endemic in Ethiopia.Buffy coat contains more parasite DNA than whole blood and plasma therefore PCR is more sensitive for buffy coat than for whole blood(33).In two dog (dog 31 and dog 19) samples, there was PCR signal in whole blood, buffy coat and plasma at least in one PCR run of repeated PCRs.However, there was still variation in ct Low DNA concentration detected in different PCR runs and different sample types make it difficult to calculate parasite load in blood and chance of the parasite to be picked up by the sand-fly.In principle, if there is at least one parasite in a sample, eluted in 50l extraction buffer from which 2l template DNA is included for PCR, it should also give positive result in PCR with 4 fg DNA (provided 100% extraction efficiency).However, with less than 1 fg DNA detected in most PCR runs of positive samples from 200l sample extract, one can extrapolate that the chance of a sand-fly picking up a parasite in a blood meal would be less than 1 in 100 times considering a sand-fly to pick 2-3l blood in one sting.Leishmania is not a true blood parasite thus more DNA would be expected to be extracted from tissue samples than blood samples by conducting a follow up study identifying DNA in skin samples which is more relevant to sandfly transmission.Though role of dogs as a reservoir for transmission was not confirmed in this study, presence of parasite DNAs in blood could be an indication that dogs could play a role in transmission as in other regions of east Africa(34,35)and India(36)where dog is considered a possible reservoir for transmission of the parasite L. donovani.
PCR results of positive dog samples show discrepancy in DNA amount (fg) in different PCR runs of same and different sample types (whole blood, buffy coat and plasma).Reasons for this variation could be: (i.), variation in the amount of parasite DNA within the 2l sample DNA included in RT-qPCR.When there are few circulating parasites or DNA in the blood, there is a small chance that they are picked up however, why variation in ct or DNA amount from the same sample in repeated PCR runs is not clear.(ii.), variation in the amount of parasite DNA in different sample types (whole blood, buffy coat and plasma).Buffy coat is the portion of blood that contains concentrates of white blood cells.se e m a n u s c r i p t D O I f o r d e t a i l s . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.It is made The copyright holder for this preprint (which this version posted April 11, 2023.; https://doi.org/10.1101/2023.04.10.536327doi: bioRxiv preprint