Comparison of virulence genes in Proteus species isolated from human and pet turtle

Abstract

The current study was aimed to investigate the prevalence of ureC, rsbA, zapA and mrpA virulence genes using polymerase chain reaction (PCR) in Proteus spp. isolated from 5 commercially popular species of pet turtles and comparison of the mrpA gene sequences of Proteus mirabilis isolates with human clinical isolates. A total of 24 isolates in pet turtles were identified, comprised of P. mirabilis (15), Proteus vulgaris (7) and Proteus hauseri (2). The prevalence of ureC, rsbA, zapA and mrpA genes among all identified Proteus spp. isolates were 91.7%, 50%, 45.8% and 45.8%, respectively. The average percentage similarities of mrpA gene sequence of pet turtle P. mirabilis isolates to human urinary and respiratory isolates were 96.35% and 94.85%, respectively. The prevalence of virulence genes and high similarity of mrpA gene sequences between pet turtles and human P. mirabilis isolates revealed that though pet turtles are healthy, these animals may pose a potential risk of urinary and respiratory infections to humans.

Introduction

The genus Proteus is a Gram-negative bacillus that belongs to the Enterobacteriaceae family. Members of the genus Proteus are widespread in the environment and the gastrointestinal tract of human and animals (Hegazy, 2016 ▶).

Proteus is known as a nosocomial, opportunistic pathogen and is more common in community-acquired infections (Omoruyia and Evangelista, 2014 ▶). Proteus mirabilis and P. vulgaris have been reported to cause wound infections, respiratory tract infections and both community-acquired and catheter-associated urinary tract infections (UTI) (Li and Mobley, 2002 ▶; Trivedi et al., 2015 ▶). In addition, Hordijk et al. (2013) ▶ have reported urinary and kidney infections of companion animals, diarrhea in cats and dogs.

Virulence of the Proteus spp. is caused by several virulent factors and these virulent factors are regulated by virulent genes encoded in operons (Manos and Belas, 2006 ▶). This study focused on zapA, rsbA, mrpA and ureC genes. Since they have been identified to code most important virulent factors and have been found to be more common in previous studies (Abbas et al., 2015 ▶; Alsherees et al., 2016). One of the prominent features of Proteus spp. is the ability to swarm on solid surfaces. Even though several genes are associated with the swarming phenomenon as cheW, gidA and cldA genes, the rsbA gene is important for the swarming regulation (Rather, 2005 ▶).

Urease is the most important enzyme for kidney and bladder stone formation in Proteus infection. The ureA, ureB, ureC, ureD, ureE, ureF, ureG and ureR genes on ure operon are responsible for the production process of urease enzyme and previous study pointed out ureC as a major gene, causative for urease production (Li and Mobley, 2002 ▶). The zap operon encoded by zapA, zapB, zapC and zapD genes is important for the production of protease, especially zapA for regulating IgA protease expression during the differentiation of swimmer cells to swarmer cells (Walker et al., 1999 ▶).

A variety of fimbriae have been detected in P. mirabilis. The vital type is MR/P fimbria encoded by mrpA, mrpB, mrpC, mrpD, mrpE, mrpF, mrpG and mrpI genes. The mrpA gene is significantly important to the pathogenicity, since it contributes numerous virulent factors such as adherence of bacteria to the epithelial tissue, biofilm formation, and swarming phenomenon (Rocha et al., 2007 ▶).

According to the comparison of genetic characteristics among bacterial isolates of human and animal origin, the issue of possible transmission risk to human has been discussed previously (Barbour et al., 2012 ▶). Nucleotide sequences of a few genes of human isolated P. mirabilis were compared with chicken and dog isolates with regard to fimbriae (Barbour et al., 2012 ▶; Harada et al., 2014 ▶).

Nowadays, pet turtles are known as a potential risk in public health, and Proteus spp. cause anorexia, pneumonia, depression, and the death of pet turtles (Henriksen, 1972 ▶). Therefore, the objectives of this study were to investigate mrpA, zapA, rsbA and ureC genes in Proteus spp. isolated from pet turtles and to compare the acquired mrpA sequences of P. mirabilis with human clinical isolates.

Materials and Methods

Fifty-two pet turtles including 2 African side-neck turtles (Pelusios castaneus), 31 Chinese stripe-necked turtles (Ocadia sinensis), 10 river cooters (Pseudemys concinna concinna), 3 Western painted turtles (Chrysemys picta belli) and 6 yellow-bellied sliders (Trachemys scripta scripta) were purchased from pet shops and online markets. The turtles were under 4 weeks of age. The turtles were raised under laboratory conditions following the general husbandry method (Bluvias and Eckert, 2010 ▶).

Isolation and identification of Proteus spp. were performed according to standard procedure (Senior, 1997 ▶; Back et al., 2016 ▶). PCR amplification of mrpA, zapA, rsbA and ureC genes was performed using specific primers and PCR conditions given in Table 1.

Table 1:

Description of gene targets and the corresponding primers used for the specific virulent genes in PCR assay

Gene	Primer type	Primer sequence 5´-3´	Size bp	T (°C) a	References
mrpA	mrpAF	ACACCTGCCCATATGGAAGATACTGGTACA	550	40°C	Barbour et al. (2012)
	mrpAR	AAGTGATGAAGCTTAGTGATGGTGATGGTGATGAGAGTAAGTCACC
rsbA	rsbAF	TTGAAGGACGCGATCAGACC	467	58°C	Abbas et al. (2015)
	rsbAR	ACTCTGCTGTCCTGTGGGTA
ureC	ureCF	GTTATTCGTGATGGTATGGG	317	56.2°C	Ali and Yousif (2015)
	ureCR	ATAAAGGTGGTTACGCCAGA
ZapA	ZapAF	ACCGCAGGAAAACATATAGCCC	540	59°C	Ali and Yousif (2015)
	ZapAR	GCGACTATCTTCCGCATAATCA

^a Annealing temperature

Randomly selected five mrpA amplicons of P. mirabilis were purified using Expin™ PCR SV kit (GeneAll^®, Korea) and sent to Cosmogenetech Co. Ltd., Daejeon, Korea for direct sequencing. The nucleotide sequences of the mrpA gene from turtles were compared with the sequences of the mrpA gene from human isolates (respiratory and urinary) previously reported by Barbour et al. (2012) ▶ and using BLAST option of NCBI (Basic Local Alignment Search Tool, BLAST v. 2.2.15, www.ncbi.nlm.nih.gov.)

All applicable international, national, and/or institutional guidelines for the care and use of animals were followed.

Results

Twenty-four isolates were positive in biochemical tests and 16S rRNA sequencing could identify them up to species level. Out of 24 bacterial isolates, 2, 7 and 15 isolates were identified as P. hauseri, P. vulgaris and P. mirabilis, respectively. The population of isolated bacteria predominantly consisted of P. mirabilis (Table 2).

Table 2.

Distribution of Proteus spp. isolated from pet turtles

Gene	Number (%) of positive strains for genes			Total positive strains (n=24)
	P. mirabilis (n=15)	P. vulgaris (n=7)	P. hauseri (n=2)
ureC	15 (100)	7 (100)	0 (0)	22 (91.7)
rsbA	12 (80)	0 (0)	0 (0)	12 (50)
zapA	11 (73.3)	0 (0)	0 (0)	11 (45.8)
mrpA	11 (73.3)	0 (0)	0 (0)	11 (45.8)

The UreC was the most prevalent gene and was identified in twenty-two isolates (91.7%). However, the ureC gene was detected only in P. mirabilis and P. vulgaris with 100% detection rate in both species (Table 3). Among four virulent genes, rsbA was detected at the ratio of 50% (12/24) while zapA and mrpA exhibited similar percentage, 45.8% (11/24). 80% of rsbA, 73.3% of zapA, and 73.3% of mrpA were observed in P. mirabilis (Fig. 1). In P. vulgaris ureC was the only gene detected and none of the genes were identified in P. hauseri isolates.

Table 3.

Distribution of virulent genes of Proteus spp. isolated from pet turtles

Turtle	Number (%) of isolates			Isolation rate of Proteus spp. (%)
	P. mirabilis	P. vulgaris	P. hauseri
Chinese stripe-necked turtle (n=31)	7 (22.6)	4 (12.9)	0 (0)	11 (35.5)
River cooter (n=10)	6 (60)	2 (20)	1 (10)	9 (90)
Yellow-bellied slider (n=6)	1 (16.7)	0 (0)	1 (16.7)	2 (33.4)
Western painted turtle (n=3)	1 (33.3)	0 (0)	0 (0)	1 (33.3)
African side-neck turtle (n=2)	0 (0)	1 (50)	0 (0)	1 (50)
Total (n=52)	15 (28.8)	7 (13.5)	2 (3.8)	24 (46.1)

Fig. 1.

Amplicons detection by PCR in P. mirabilis and P. vulgaris of pet turtle isolates (Lanes 4-25). a: mrpA gene, b: rsbA gene, c: ureC gene, and d: zapA gene. Other lanes are for the negative control (Lane 3), P. mirabilis reference isolate (American Type Culture Collection) (Lane 2), and genetic base-pair ladder (Lane 1

Mean similarity percentages with regard to mrpA gene of pet turtles and four human urinary P. mirabilis isolates were 96.2%, 96.4%, 97.0%, and 95.8%, respectively. Besides, human respiratory and turtle isolates showed 94.4%, 95.4%, 94.6%, and 95.0% of similarity percentages. Mean of means was calculated using similarity percentages and it was 96.4% in urinary isolates and 94.9% in respiratory isolates (Table 4).

Table 4.

Average similarity percentages of mrpA sequences of P. mirabilis recovered from pet turtles in comparison to human respiratory and urinary clinical isolates

Turtle isolates	Mean % similarity in mrpA amplicon sequences of pet turtles to human isolates
	Human urinary isolates				Mean	Human respiratory isolates				Mean
	1	2	3	4		1	2	3	4
1	98.0	98.0	98.0	98.0	98.0	96.0	98.0	96.0	98.0	97.0
2	95.0	95.0	95.0	95.0	95.0	93.0	95.0	94.0	95.0	94.3
3	97.0	96.0	98.0	96.0	96.8	96.0	96.0	95.0	94.0	95.3
4	98.0	98.0	96.0	95.0	96.8	93.0	95.0	93.0	95.0	94.0
5	93.0	95.0	98.0	95.0	95.3	94.0	93.0	95.0	93.0	93.8
Mean	96.2	96.4	97.0	95.8		94.4	95.4	94.6	95.0
Mean of means					96.4*					94.9**

^{*, **} Mean of means in a row followed by different symbols are significantly different (P<0.05)

Discussion

Proteus spp. was detected in many wild and domestic animals particularly with no clinical signs (Drzewiecka, 2016 ▶). Since Proteus spp. is an opportunistic pathogen, even healthy turtles can be asymptomatic carriers of the pathogen (Hossain et al., 2016 ▶). In the present study, P. mirabilis was the predominant species followed by P. vulgaris as 62.5% (15/24) of P. mirabilis, 29.17% (7/24) of P. vulgaris and 8.34% (2/24) of P. hauseri. Dalia (2015) ▶ has reported the same results with clinical isolates.

The current study was mainly focused on the detection of virulence genes. Several virulence determinants contribute to the pathogenicity of Proteus spp. (Schaffer and Pearson, 2015 ▶). The ureC gene is responsible for the elevation of pH of urine and resulting in stone formation (Mohammed et al., 2014 ▶). In this study, ureC amplification of P. mirabilis and P. vulgaris was 100%, which reveals a higher frequency of ureC compared with the rest of the genes. It indicates the crucial role of ureC for the virulence of the Proteus spp. Regarding ureC gene, these results are in agreement with a previous study, indicating the presence of ureC approximately 96.6% among isolates obtained from patients with UTI (Ali and Yousif, 2015 ▶).

The rsbA gene could not be amplified in P. vulgaris and P. hauseri. However, the majority of positive strains were P. mirabilis and it is in line with the previous studies (Rather, 2005 ▶; Abbas et al., 2015 ▶). In the current study, a phenotypic character of swarming, the bull’s-eye ring was exhibited by all the isolates. Nevertheless, swarming regulated genes are not necessarily required for swarming, for the reason that so many genes and operons are involved in the process (Pearson et al., 2010 ▶).

The zapA gene which is coding for protease enzyme production and mrpA gene which is mainly important for fimbriae gene were identified in similar number of isolates 11 (45.8%). The zapA and mrpA could not be detected in P. vulgaris and P. hauseri isolates. Nevertheless, the frequency of both genes in P. mirabilis was 73.3% in disagreement with previous studies which were 30% (Alsherees et al., 2016) and 100% (Ali and Yousif, 2015 ▶) in clinical P. mirabilis isolates. Although the prevalence of mrpA gene in previous studies was small in percentage values, the most important gene among tested gene is mrpA due to its contribution to several virulent factors (Abbas et al., 2015 ▶). Therefore, mrpA was the selected gene for comparison of gene sequences.

Furthermore, comparison of mrpA amplicon sequences of pet turtles to human isolates revealed similarity percentages. Fimbriae are imperative for several virulent factors of Proteus species. Among 6 fimbriae types and 17 fimbrial operons, MR/Pis the most important fimbria (Różalski et al., 2012 ▶). This type of fimbria is encoded by mrp operon containing 10 genes located on the bacterial chromosome. The main structural subunit of these fimbriae is the MrpA protein. Though other fimbriae are present only in P. mirabilis, the mrp operon was shown not only in P. mirabilis strains but also in other species of the Proteus genus. In our results, comparison of pet turtles to human urinary isolates, using mrpA gene of P. mirabilis represents 96.40% of mean similarity and pet turtles to human respiratory isolates using the same gene of P. mirabilis represents 94.90% of mean similarity.

Previous studies proved that direct contact even with healthy turtles can transmit diseases to humans and if people, especially immune suppressed adults and children, have direct physical contact with turtles and aquarium water, there might be a potential risk to transmit pathogens and to cause urinary infections (Mermin et al., 2004 ▶; Hidalgo-Vila et al., 2007 ▶).

In conclusion, the high prevalence of Proteus species, especially P. mirabilis in pet turtles, and their similarity of virulence gene characteristics with human clinical isolates could be emphasized. Thus the potential health risk of infection of Proteus spp. from pet turtles should not be underestimated. Expanding this research further with more virulent genes could be important for a better understanding of the pathogenicity of Proteus spp. harbored by pet turtles. According to our knowledge, this is the first report designed to detect zapA, rsbA, mrpA and ureC genes in Proteus isolates of pet turtles.

Conflict of interest

The authors declare no conflict of interests.