In-silico identification of potential antigenic proteins in Bartonella bacilliformis for the serological diagnosis of Carrions’ disease

The current methods for inferring antigenic, pathogenic or virulence factors in bacteria include in-silico or computational tools. Bartonella bacilliformis, the causal agent of Carriońs disease (CD) is a Gram-negative microorganism transmitted by the sand fly Lutzomya verrucarum mainly in Peruvian Inter-Andean valleys. A better understanding of the pathogenicity of B. bacilliformis, the implementation of serological diagnostic methods and the development of candidate vaccines for the control of CD could be facilitated by the identification of outer membrane exposed proteins such as outer-membrane beta-barrels (OMBB) and outer-membrane lipoproteins (OMLPP). In this study we present the in-silico identification of OMBBs and OMLPPs in B. bacilliformis. The present analysis identified 32 OMBBs and 9 OMLPPs, of which 15 and 4 are reported for the first time as potential antigenic, pathogenic and virulence-related proteins, respectively. Future implications of this study are discussed and compared with proteins validated by experimental assays.


25
Carrion's disease (CD) is an endemic illness of Andean countries such as Peru and Ecuador, with an 26 additional historical presence in Colombia [1,2]. In Peru, CD represents one of the main challenges 27 for public health due to poverty and poor sanitation in endemic localities, affecting children with 28 chronic malnutrition. Bartonella bacilliformis, the etiological agent of CD, is a Gram-negative 29 pleomorphic microorganism transmitted by sand flies of the Lutzomyia genus, especially L. 30 verrucarum [2]. Climate change and variability in inter-Andean valley rainfall associated with the 31 El Niño-Southern oscillation (ENSO) favors the reproduction of L. verrucarum and the consequent 32 emergence of local CD outbreaks with a non-negligible number of cases [3]. CD clinical 33 manifestations are diverse as the microorganism parasites human erythrocytes generating an 34 "acute phase, called Oroya fever, characterized by both anemia and febrile illness [2]. 35 Nevertheless, the nature of the initial symptoms may be confused with that of other infectious diseases, such as malaria, dengue or others. This is of special relevance because the absence or 37 delay of adequate treatment may result in fatal outcomes [2]. In this sense, mortality rates among 38 untreated or inadequately treated patients of up to 88% have been described [4]. Of note,in Peru,39 the overall Oroya fever lethality ranges between 0.5 and 3%, with about 10% of severe cases 40 attending reference hospitals having a fatal outcome [2].

41
Eventually, after a period of several weeks to months following the acute phase, a non-life 42 threating eruptive or "verrucose phase" distinguished by notable skin eruptions develops and may 43 be presented in the absence of a previous acute infection [1,2].

44
In addition to the clinical manifestations described above, patients with asymptomatic bacteremia 45 have been identified and are considered as potential reservoirs of Bartonella bacilliformis and a 46 potential source of infection for susceptible persons [3].

47
While bacteriological cultures are accurate, they are time-consuming given the slow growth rate of 48 B. bacilliformis (about 4-15 days), and therefore have limited usefulness for diagnostic purposes. 49 On the other hand, blood smear detection is fast, accurate and low cost, and is thereby largely 50 used for diagnostic purposes, especially in rural areas. Nevertheless, it has a low sensitivity, 51 generating false negatives, limiting the confirmation of any diagnosis [5,6]. Molecular methods 52 based on the detection of specific gene regions such as gltA, ribC and ialB have been proposed, 53 being either limited to gene amplification or combined with a subsequent process, such as 54 enzymatic digestion [7]. Notwithstanding, in addition to the difficulties in implementing these 55 techniques in most rural areas affected, some of these genes are not specific for Bartonella 56 bacilliformis, causing cross reactivity with other pathogens [7]. It should be taken into account that 57 the current serological techniques available are in-house with no commercial development, having 58 a limited specificity, and potential cross reactivity with other Bartonellaceae or other 59 microorganisms cannot be ruled out, similar to what has been reported for different 60 immunological approaches for the detection of Bartonella quintana and Bartonella henselae 61 infections [8,9]. Regarding rapid diagnostic tools that can be used in endemic areas, different 62 antigenic candidates have been proposed but none has been introduced in clinical practice 63 [10,11].

64
Data about immune response to B. bacilliformis is limited, with antibody immunity build-up for the 65 development of partial immunity being described considering immunoglobulin M (IgM) as a 66 biomarker of the acute phase and immunoglobulin G (IgG) as a marker of previous exposure 67 [10,12].

68
The finding of specific outer membrane proteins (OMPs) of B. bacilliformis could establish the 69 basis for the future implementation of an accurate and sensitive serological assay for the diagnosis 70 of CD as OMPs activate immune response and participate in virulence mediating pathogen-host 71 interactions [13]. In Gram-negative bacteria, the tertiary structure of OMPs includes beta-barrel 72 structure composed of a variable number of beta-strands. Another type of exposed OMPs known 73 as outer membrane lipoproteins are also involved [14,15]. In addition, both outer-membrane 74 beta-barrels (OMBB) and outer-membrane lipoproteins ( antigenic re-activity based on scientific studies were found in the PubMed repository.

111
Lipoproteins with the following three characteristics were selected: 1) containing signal-peptide 112 potentially cleaved by signal peptidase II (SPaseII) in 3 out of the 5 tools, 2) located on the outer 113 membrane in 3 out of 4 tools, and 3) recognized as lipoproteins by 3 out of the 5 tools. Beta-barrel 114 proteins with the following two characteristics were selected: 1) located in the outer membrane in 115 3 out of 4 tools and 2) recognized as beta-barrel by 3 out of 4 predictive tools.

132
The present study aimed at finding potential antigenic OMPs and the identification of potential 133 antigenic domains in proteins with known immunological reactivity in B. bacilliformis. By 134 contrasting subcellular localizations, the presence of signal peptides, lipoprotein and beta-barrel 135 protein predictors followed by proteome-proteome comparisons between human, murine, febrile-136 associated and other Bartonella species, 32 OMBB proteins and 11 OMLPs were identified, which 137 are shown in Tables 1 and 2. All 9 experimental proteins with immunological reactivity followed 138 the same procedure (Table 3). For the present descriptions, selected proteins were listed in terms 139 of accession numbers followed by their order in the annotated genome (GenBank accession: 140 NZ_CP045671.1).  the beta-barrels, we did not find protein homologs in human, mouse or febrile illness-related 167 species proteomes (Figures 2 and 3). 3D modelling corroborated the beta-barrel structures in 168 these proteins (Figure 4). are colors as beta-strand (green), alpha-helix (blue) and coil (orange). Secondary structures 183 show beta-strand (green), alfa-helix (blue) and coil (orange). FlgA, FlgH, FliF and FlgI had percent similarities between 75% and 84% and percent coverage 203 between 86% and 97%. All flagellar beta-barrels presented similarities with febrile illness-204 associated species proteins with percent identities between 30% and 39% and percent 205 coverages between 55% and 91% ( Figure 3).  TamB  209 presented mean percent similarities between 42% and 59%, with percent coverages above 210 84% compared with other Bartonella, whereas the mean percent similarities were less than 211 35% compared with febrile illness-associated species proteomes. Signal peptidase I showed a 212 mean percent similarity of 75% and a mean percent coverage of 99% ( Figure 3). BamA presented percent identities between 45% and 50% and percent coverage greater than 218 82% compared with other Bartonella, while percent identities were between 28% and 33% 219 with percent coverage greater than 75% compared with febrile illness-associated species 220 proteomes (Figure 3). They also presented percent similarities between 23% and 26% and 221 percent coverages between 53% and 62% compared with human and mouse proteomes 222 (Figure 2).

223
Outer membrane lipoproteins 224 The present methodology identified 9 outer membrane lipoproteins. Only one lipoprotein "Do 225 family serine endopeptidase" (WP_005767379.1, 769) presented mean percent similarities and mean percent coverages close to 35% with human and mouse proteins ( Figure 5). The 3D 227 structures are shown in Figure 6.  1, 769). The first two proteins presented a mean percent identity between 245 32% and 34%, a mean percent coverage between 36% and 100% in comparison with febrile 246 illness-associated species proteomes, and a mean percent identity between 44% and 51%, and 247 a mean percent coverage between 65% and 100% in comparison with Bartonella species 248 proteomes. As described above, only Do family serine endopeptidase presented similarities 249 with human and mouse proteins ( Figure 5).

298
In this study we identified OMBB and OMLP of B. bacilliformis. Some of these proteins have 299 been identified and or described by other researchers while others have not previously been 300 reported. To achieve the objectives of the study, different OMBB predictors, lipoprotein 301 predictors, subcellular localization predictors and signal-peptides predictors were used. In-302 silico identification of candidate antigenic proteins has been used to propose vaccine-303 candidates and in the search for antibody targets in the development of serological tests. B.
bacilliformis outer-membrane and extracellular proteins have been analyzed by the 305 identification of epitopes in the design of an artificial multi-epitope protein with positive 306 results in murine model [37]. In a recently published article, Dichter et al. [38] tested the 307 immunogenicity of 22 candidate B. bacilliformis proteins identified by a reverse vaccinology 308 approach using Vaxign-1 software. The present study identified 19 exact or homologs of these 309 proteins.

310
Outer membrane beta-barrel proteins 311 Autotransporter outer-membrane beta-barrel proteins (ATOMP). YadA-like. Nevertheless, Vomp could not be modeled by Phyre2. These proteins contain 7 and 324 2 "coiled stalk" domains, respectively, with low percent identities and coverages compared 325 with febrile and other Bartonella species.  (or greater  349 than) 78% compared with other Bartonella species limit the use of these proteins as 350 discriminant candidates but not as vaccine or generic targets. Membrane protein assembly/Translocases. The LptD protein is involved in the 377 assembly of polysaccharides in the outer membrane [69]. Mass spectrometry analysis 378 conducted by Li et al. (2014)