• No se han encontrado resultados

Virulence evaluation of Listeria monocytogenes strains isolated in Colombia

N/A
N/A
Protected

Academic year: 2020

Share "Virulence evaluation of Listeria monocytogenes strains isolated in Colombia"

Copied!
21
0
0

Texto completo

(1)

Virulence evaluation of Listeria monocytogenes strains isolated in Colombia Angélica Garzón1, María Consuelo Vanegas2, María Mercedes Zambrano3

1,2: Laboratory of Food Microbial ecology - LEMA, Biological Sciences Department, Universidad de los Andes, Colombia. 3: Corpogen, Research and Biotechnology, Colombia.

Introduction

Listeria monocytogenes is a zoonotic and intracellular pathogen ubiquitously distributed in the environment, especially in plant matter and soil. The principal reservoirs of Listeria

are soil, forage, water and infected animals. 99% of listeriosis cases are foodborne, through the consumption of milk and dairy products, artisan cheeses, beef, chicken, pork and ready to eat foods. Severe illness mainly occurs in pregnant women, the unborn child, infants, the elderly and those with compromised immune systems, withmortality rates ranging between 20% and 30% [1–13].Illness caused by the consumption of contaminated food has a wide economic and public health impact worldwide and multiple efforts have been made in the study of virulence genes of this food borne pathogen[14–16].

In Colombia, notification of this pathogen is not mandatory. However, the estimated prevalence is of 19% for L. monocytogenes according to UERIA (Risk Assessment Unit for Food Safety in Colombia)[4]. Also by 2010, 158 isolates from clinical patients were obtained by the National Health Institute, in agreement with the presence of this microorganism in both food and patients[4].

INVIMA (National Institute for Drug and Food Surveillance in Colombia) established that 57.6% of L. monocytogenes isolated from food between 2000 and 2009 belonged to 4b serotype which it is known as a virulent serotype [17]. However, there is little information regarding the characteristics of Colombian strains, particular with respect to their virulence. Therefore, research on the molecular determinants and virulence properties of Colombian isolates would contribute greatly to understanding the types of strains circulating in various environments and their ability to cause listeriosis in humans.

Currently, L. monocytogenes can be classified into 13 serotypes (1/2a,1/2b, 1/2c, 3a, 3b, 3c, 4a, 4ab, 4b, 4c, 4d, 4e, 7) according to the flagellar and somatic antigens. The majority of human listeriosis outbreaks worldwide have been linked to what have been described as virulent serotypes 4b, 1/2b and 1/2a. An association between L. monocytogenes flagellar antigens and distinct partial DNA sequences of the listeriolysin gene (hly) has been described. Remaining serotypes are often found on different environmental niches as food, natural environments or ruminats in some cases [17, 18].

There have been extensive studies on the molecular mechanisms involved in the pathogenesis of L. monocytogenes in an effort to understand the disease and propose

(2)

alternatives for control [9, 19, 20]. L. monocytogenes expresses virulence factors directly related with the process of infection and its intracellular lifestyle, most of them located in the LIPI-1 (Listerial Pathogenicity Island 1). Those include hly which encodes a pore forming toxin Listeriolysin O; actA which encodes an actin polimerization protein involved in cell to cell motility and the prfA gene encoding a 27 KDa protein, which regulates the expression of L. monocytogenes virulence genes. PrfA also controls the expression of inlA and inlB, virulence genes, which code for internalin proteins and are not located in LIPI-1 (Figure 1). Internalins A and B are also necessary during the infectious process since they mediate the entry to the host cell and can induce phagocytosis even in non-phagocytic host cells[20–25].

Figure 1: Virulence genes regulated by PrfA[22].

Regulation by PrfA, a central regulator of virulence gene expression in L. monocytogenes, has been studied in great detail [21, 22, 25–27]. The primary mechanism for PrfA regulation of virulence gene expression is by direct activation of transcription at promoters bearing a palindromic DNA sequence, the PrfA box

(consensus tTAACanntGTtAa) located at -45 bp. This sequence is key for interaction with the DNA binding motif of PrfA. Promotors with perfectly symmetric prfA boxes are activated more efficiently by PrfA. Differences in PrfA affinity are present due to the type of palindromes in prfA boxes. Functional prfA boxes have maximum two and exceptionally three mismatches[21–23, 28, 29].

PrfA is a member of the Crp/Fnr family of bacterial transcription factors[20–22].PrfA and the enterobacterial regulator Crp (cAMP receptor protein) are homodimers with very similar subunits including the motif for union to DNA. This family of bacterial transcription factors normally requires a cofactor for their activation. In Crp, the binding of cAMP to the N-terminal domain causes the protein, which is inactive without its ligand, to adopt a conformation with a higher affinity for the target DNA. Structure analysis of Crp

(3)

indicates the presence of a tunnel where cAMP is bind causing a conformational change and activation of Crp regulator. Site directed mutagenesis of Crp gives evidence that this conformational change activates the regulator without its cofactor cAMP.

PrfA can be found in two different states: inactive outside the host cell and active after a protein conformational change produced intracellulary. These structural functional analogies with Crp support the notion that PrfA activity is also regulated allosterically [20–22, 30]. There is also evidence of the presence of a tunnel in PrfA similar to the Crp union tunnel to cAMP. However, the cofactor for PrfA is unknown in L. monocytogenes

and it has been proposed that it can be related to carbon source metabolism in host cells[20, 22]. Structure similarity of PrfA and Crp could help to understand the protein functionality of PrfA in L. monocytogenes Colombian isolates.

Thus, there is a great amount of information about L. monocytogenes virulence mechanisms, but little is known regarding many of these aspects in strains found in our country. The aim of this research was therefore to analyze L. monocytogenes strains isolated in Colombia in order to assess their characteristics and virulence capacity. This will be done by analyzing molecular determinants associated with virulence, looking at the PrfA structure to understand possible interactions with prfA boxes, and carrying out

in vivo cell assays.

Materials and Methods Bacterial Strains

10 L. monocytogenes strains belonging to the LEMA (Laboratorio de Ecología Microbiana y de Alimentos) collection maintained at -20°C in 20% glycerol, were reconstituted in Brain Infusion Heart Broth (BHI) and both virulent and non-virulent serotypes were selected (Table 1). Strains were maintained in Standard Plate Count Agar at 4 °C. For analysis, they were cultured for 18 h in BHI broth at 37 °C. L. monocytogenes ATCC 7644 was used as PCR positive control and L. ivanovii ATCC 19119 and L. innocua ATCC 73016 were used as negative controls for all performed assays.

Code Strain Origin Serotype

LMO175 L. monocytogenes Isolated from Patient Blood culture 4b*

LMO179 L. monocytogenes Isolated from Patient Cerebrospinal fluid 4b*

LMO135 L. monocytogenes Isolated from sausage 4b*

(4)

LMO69 L. monocytogenes Isolated from milk 4b*

LMO206 L. monocytogenes Isolated from cow´s liver 4b*

LMO195 L. monocytogenes Isolated from stool sample 1/2a *

LMO86 L. monocytogenes Isolated from siphon floor 4a

LMO418 L. monocytogenes Isolated from cheese Unknown ATCC

7644 L. monocytogenes Reference ATCC Strain 7644 1/2c ATCC

19119 L. ivanovii Reference ATCC Strain 19119 ATCC

73016 L. innocua Reference ATCC Strain 73016

Table 1: Listeria sp. strains used in this study. Serotypes associated with virulence are indicated with an asterisk (*).

in silico analysis of Listeria monocytogenes Pathogenicity Island (LIPI-1)

In order to determine if there was variability in the prfA boxes of L. monocytogenes

strains isolated in Colombia, an analysis of LIPI-1 was first performed using L.

monocytogenes genomes downloaded from NCBI RefSeq database. For this, 24

complete genomes of L. monocytogenes were downloaded, 11 of them correspond to virulent serotypes (4b, 1/2a and 1/2b), 11 to non-virulent serotypes (3a, 3b, 3c, 1/2c, 4a, 4c, 4d, 4e and 7) and 2 without previously known serotype (Table 2). In addition, virulence gene sequences (prfA, plcA, mpl, hly, actA e inlA) and complete LIPI-1 clusters were also downloaded from NCBI and these sequences were identified in each complete genome using Blast.

Strain Serotype Accesion number Source

L. monocytogenes

ATCC 19117 4d NC_018584.1

Comparative genomics of the bacterial genus Listeria: Genome evolution is characterized by limited gene acquisition and limited gene

loss. Unpublished article. L. monocytogenes

L99 4a NC_017529.1 Variability in sequence, expression and immune-evasion in pathogenic Listeriae. Unpublished article.

L. monocytogenes

085578 1/2a * NC_013766.1

High-throughput genome sequencing of two Listeria monocytogenes clinical isolates during a large foodborne outbreak. BMC Genomics 11,

120 (2010). Strain isolated from infected patient in Canada 2008 outbreak trough consumption of contaminated meat.

L. monocytogenes

F2365 4b * NC_002973.6

Whole genome comparisons of serotype 4b and 1/2a strains of the foodborne pathogen L. monocytogenes reveal new insights into the

core genome components of this species. Nucleic Acids Res. 32 (8), 2386-2395 (2004). Strain isolated from Jalisco cheese in 1985 California

outbreak L. monocytogenes

M7 4a NC_017537.1

Genome Sequence of the Nonpathogenic Listeria monocytogenes Serovar 4a Strain M7. J. Bacteriol. 193 (18), 5019-5020 (2011). Strain

(5)

L. monocytogenes

085923 1/2a * NC_013768.1

High-throughput genome sequencing of two Listeria monocytogenes clinical isolates during a large foodborne outbreak. BMC Genomics 11,

120 (2010). Strain isolated from infected patient in Canada 2008 outbreak trough consumption of contaminated meat. L. monocytogenes

Clip80549 4b * NC_012488.1 Variability in sequence, expression and immune-evasion in pathogenic Listeriae. Unpublished article. L. monocytogenes

Finland 1998 Unknown NC_017547.1 The Genome Sequence of Unpublished article. Listeria monocytogenes strain Finland 1998. L. monocytogenes

J0161 1/2a * NC_017545.1 article. Strain isolated from a deli in an epidemic outbreak in USA in 2000. The Genome Sequence of Listeria monocytogenes J0161. Unpublished L. monocytogenes

HCC23 4a NC_011660.1 Genome Sequence of Lineage III J. Bacteriol. 193 (14), 3679-3680 (2011). Isolated from healthy catfish. Listeria monocytogenes Strain HCC23. L. monocytogenes

EDGe 1/2a * NC_003210.1 Comparative and functional genomics of (5543), 849-852 (2001). Listeria spp. Science 294 L. monocytogenes

SLCC2482 7 NC_018591.1 The genome sequences of 11 strains of covering all known serotypes. Unpublished article Listeria monocytogenes, L. monocytogenes

SCCL2378 4e NC_018585.1 The genome sequences of 11 strains of covering all known serotypes. Unpublished article Listeria monocytogenes, L. monocytogenes

SCCL2479 3c NC_018589.1 The genome sequences of 11 strains of covering all known serotypes. Unpublished article Listeria monocytogenes, L. monocytogenes

SCCL2755 1/2b * NC_018587.1 The genome sequences of 11 strains of covering all known serotypes. Unpublished article Listeria monocytogenes, L. monocytogenes

SCCL5850 1/2a * NC_018592.1 The genome sequences of 11 strains of covering all known serotypes. Unpublished article Listeria monocytogenes, L. monocytogenes

L312 4b * NC_018642.1 The genome sequences of 11 strains of covering all known serotypes. Unpublished article Listeria monocytogenes, L. monocytogenes

SCCL7179 3a NC_018593.1 The genome sequences of 11 strains of covering all known serotypes. Unpublished article Listeria monocytogenes, L. monocytogenes

SCCL2376 4c NC_018590.1 The genome sequences of 11 strains of covering all known serotypes. Unpublished article Listeria monocytogenes, L. monocytogenes

SCCL2372 1/2c NC_018588.1 The genome sequences of 11 strains of covering all known serotypes. Unpublished article Listeria monocytogenes,

L. monocytogenes

07PF0776 4b * NC_017728.1

Genome sequence of Listeria monocytogenes 07PF0776, a cardiotropic serovar 4b strain. J. Bacteriol. 194 (13), 3552 (2012). Strain isolated from

heart abscess of infected patient. L. monocytogenes

FSLR2561 Unknown NC_017546.1 The Genome Sequence of Unpublished article. Listeria monocytogenes strain FSL R2-561. L. monocytogenes

10403S 1/2a *ª NC_017544.1 The Genome Sequence of Unpublished article. Resistant to streptomicin. Listeria monocytogenes strain 10403S. L. monocytogenes

SCCL2540 3b NC_018586.1 The genome sequences of 11 strains of covering all known serotypes. Unpublished article Listeria monocytogenes, Table 2: RefSeq L. monocytogenes strains used in this study. Virulent serotypes are

indicated with an asterisk (*).

The position, length and organization of LIPI-1 (Figure 1) were determined for all downloaded genomes using bioinformatic tools as MegaBlast[31, 32], MobylePasteur[33] and T-coffee multiple alignment in order to identify initial and final positions of each one of the virulence genes tested as well as similarity [34]. Virulence genes located in the island were identified, including prfA, plcA, hly, mpl and actA. The

(6)

positions and sequences of prfA boxes were also identified for three of the LIPI-1 genes (plcA, mpl and actA), as well as for the inlA gene that is outside this island.

Bioinformatics analyses were also performed with 2 complete genomes of L. ivanovii

(Accession Numbers: NZ_CM001050.1 and NC_016011.1) and 3 complete genomes of L.

innocua (Accession numbers: NC_003212.1, NZ_CM001048.1 and NZ_CM001049.1)

downloaded from NCBI database RefSeq.

Virulence gene characterization of L. monocytogenes strains isolated in Colombia

The hly, actA and prfA virulence genes and the intergenic regions (hly – mpl; Lmo02648 – InlA) were amplified by PCR. Intergenic regions and the prfA gene were sequenced at Macrogen Inc. Primer design for amplification of intergenic regions were performed with Oligo 7 Software. PCR primers and conditions are summarized in Table 3.

Gene or

region direction Primer Sequence (5´- 3´) denaturation Initial (No. of cycles) Amplification Final hold Product Size Reference

prfA Forward CGTTTGGAGCTCTTCTTGGTGAAGCA 95°C, 5 min 95°C, 15 s; 60°C, 30 s; 72°C, 90 s (30)

72°C,

10 min 1050 bp [35] Reverse AGCAACCTCGGTACCATATACTAACT

hly Forward CGGAGGTTCCGCAAAAGATG 95°C, 5 min 95°C, 90 s; 55°C, 60 s; 72°C, 90 s (35)

72°C, 7

min 234 bp [36] Reverse CCTCCAGAGTGATCGATGTT

actA Forward TAGCGTATCACGAGGAGG 94°C, 2 min 30 s 94°C, 3 min; 53°C, 1 min; 72°C, 2 min (40)

72°C, 5

min 827 bp [37] Reverse TTTTGAATTTCATATCATTCACC

Intergenic region hly–

mpl

Forward CACTACGCTTTATCCGAAATA 94°C, 4 min 59,3°C, 40 s; 94°C, 50 s; 72°C, 30 s (37)

72°C,

4min 380 bp in this study Designed Reverse CCTGAAAAGCTATTACCATGA

Intergenic region lmo02648-

inlA

Forward GTATCACACCAGACCGAAT

94°C, 4 min 94°C, 50 s; 60°C, 40 s; 72°C, 40 s (37)

72°C,

4min 810 bp in this study Designed Reverse GCTGCCAAATACTAATATTGCT

Table 3:PCR conditions for genes amplified in this study

PrfA Modeling and Analysis

8 prfA sequences were translated using the EMBOSS Transeq tool from the European Bioinformatics Institute web site http://www.ebi.ac.uk/. Domains for each one of the translated sequences were identified using the InterProScan database[38]. Secondary structure was assessed by the Quick2D software available at the Bioinformatics Toolkit from the Max-Planck Institute for Developmental Biology available at http://toolkit.tuebingen.mpg.de/. Tertiary structure was predicted by SwissModel software and verified with the Ramachandran Plot performed using the Rampage web site available at http://mordred.bioc.cam.ac.uk/~rapper/rampage.php[39]. Three-dimensional analyses were perfomed with Patch Dock[40, 41], Swiss PdbViewer[42, 43] and Chimera Software[44].

(7)

Protein Data Bank PrfA (PDB code 2BEO) and Crp (PDB code 3RDI) proteins were downloaded for analysis and comparison with L. monocytogenes predicted PrfA proteins for strains isolated in Colombia.

In vivo virulence assay

L. monocytogenes strains and inoculum standardization

Growth curves were developed for selected strains of L. monocytogenes (Table 1) by measuring the optical density (OD) at 540 nm every hour for 12 h, when grown at 37 °C in Brain Heart Infusion broth in comparison with counts determined by plating serial dilutions on Palcam Agar Base medium. Mid-log-phase bacterial cultures were adjusted to10 ×106CFU/ml[45–48] for use in virulence assays.

Virulence assay

Human HT29 cells (Human colon adenocarcinoma grade II cell line) were grown in 24-well tissue culture plates for 48 h at 37 °C with 5% CO2 to reach a monolayer. 1 ml of each bacterial culture (Table 1), previously adjusted to approximately 10 ×106 CFU/ml, was added to each well of the 24 well plate. After 1 h of incubation at 37 °C, the HT29 cells were washed twice with 1 ml saline water (0.9% NaCl, pH 7.2) to remove free bacteria. One ml of 0.1% TritonX-100 was added to each well to loosen and lyse the adhered HT29 cells and the mixture was diluted and plated onto Blood Agar and Palcam Agar Base (Oxoid) to determine the number of adhered bacteria.

The number of bacteria invading the cells was determined as follows: After 1 h of infection, the monolayer was washed twice in saline water and extracellular bacteria were killed by incubation in DMEM (Dulbecco's Modified Eagle Medium, Life Technologies) with 100 μg/ml gentamicin for 1 h at 37 °C. The medium with gentamicin was removed and DMEM was added to each well. The plates were incubated for 2 h and 3.5 h at 37 °C. At each time interval DMEM was removed and cells were lysed with 1 ml 0.1% TritonX-100, and the number of intracellular bacteria was determined by plating serial dilutions onto Blood Agar and Palcam Agar plates. All experiments were carried out in duplicate[45–48].

Discussion and Results

in silico analysis of Listeria monocytogenes Pathogenicity Island (LIPI-1)

The LIPI-1 island of L. monocytogenes is known to contain various genes involved in virulence, however it is not evident if this genomic region in strains isolated in Colombia is similar to that present in known reference strains. As a first step towards this

(8)

characterization, and because the LIPI island is very large, we started by analyzing possible variability present in sequenced strains. To do this, the LIPI-1 genomic region was obtained from RefSeq L. monocytogenes strains found in databases and analyzed using bioinformatics tools. Completed genomes were aligned with downloaded virulence genes and LIPI-1 clusters to determine LIPI length and virulence gene positions. The length of the LIPI Island varied from 8151 to 8644 bp, a variation that was not related to the serotype of analyzed strains. However, all virulence genes were found in LIPI-1, in spite of its length. The differences in length were due to the presence of open reading frames at the end of the island, which do not encode virulence associated genes and varies between strains. This shows that the LIPI-1 island is quite conserved in terms of gene content.

An analysis of the prfA boxes in the intergenic regions between genes plcA - hly, hly – mpl, mpl – actA and upstream inlA was next carried out in order to establish if there was any variability in prfA boxes. The prfA boxes were found in two intergenic regions plcA – hly and mpl – actA, in all 24 L. monocytogenes islands analyzed from these sequenced strains. In the intergenic region hly – mpl and the region upstream of inlA, prfA boxes

were only found in 23 strains. They were not found in L. monocytogenes HCC23 non-virulent serotype 4a, indicating that there might be a possible relation between the presence of the prfA boxes, serotype and subsequently the virulence of a particular L. monocytogenes strain.

In this order of ideas, the intergenic regions that showed variability, hly – mpl and the region upstream of inlA, were chosen for analysis in Colombian L. monocytogenes

isolates. This was done to determine if they were present and if there was any variability with respect to sequenced strains or that could be correlated with strain virulence. These regions were therefore PCR-amplified using specific primers (Table 3) and sequenced. The intergenic regions plcA – hly and mpl – actA were not amplified because no variation of prfA boxes was observed in RefSeq strains, indicating a low probability of identifying any variance in our Colombian isolates.

To complement the analysis carried out in silico with L. monocytogenes strains, the presence of virulence determinants was also analyzed using genomes of the closely related microorganisms, two of L. ivanovii and three of L. innocua. L. ivanovii is associated to animal listeriosis and can cause listeriosis in humans. Complete LIPI-1 has been reported to this microorganism. On the other hand L. innocua is consider as a non-pathogenic strain and may or may not have virulence genes.

Only one strain of L. ivanovii had the prfA gene, which had an identity percentage of 86% at the nucleotide level with the L. monocytogenes prfA gene. The other virulence genes (plcA, hly, mpl and actA) of L. monocytogenes were not found in either of the L. ivanovii strains. This is contradictory with previous reports that indicate that L. ivanovii has the complete LIPI-1 and cause listeriosis in humans in sporadic cases[49].

(9)

The two strains of L. innocua contained the prfA gene, with an identity percentage of 93% with the L. monocytogenes prfA. One of the strains presented all L. monocytogenes

virulence genes with identity percentages between 91 and 97%. According to literature, L. innocua causes infections only in animals in very rare cases and may or may not have virulence genes. This is concordant with the observed results in this study[49].

Virulence gene characterization of L. monocytogenes strains isolated in Colombia. To characterize the virulence potential of L. monocytogenes Colombian isolates, the presence of hly, actA and prfA virulence genes as well as of intergenic regions identified to be variable by in silico analysis, were analyzed in Colombian L. monocytogenes

strains (Table 1). These strains were selected because they varied in terms of origin (clinical isolates, food and environmental strains) and serotype. In addition, PrfA sequences were also analyzed for their three-dimensional structure to see if they were similar to PrfAwt crystallographic structure downloaded from PDB (Protein Data Bank) and to see if cofactor tunnel reported for PrfAwt was observed in Colombian isolates. Finally, strains were analyzed for their capacity to invade HT29 cells using an in vivo cell assay.

All three virulence-associated genes (hly, actA and prfA) were detected in 50% of L. monocytogenes strains isolated from food and industrial surfaces, including the reference strain L. monocytogenes ATCC 7644 (Table 4). Absence of virulence genes in some Colombian isolates shows variation in LIPI-1 organization that could be related to the expression of virulence of this strains. The hly gene alone, however, was amplified in all tested strains, showing the potential of using this as a genetic marker for the molecular identification of L. monocytogenes. It is known that hly allows the bacterium to escape from the host cell phagocytic vacuole, a characteristic that is associated with virulence[20, 50, 51].

 

PCR profile of virulence

associated genes virulence In vivo assay on HT 29 cell

line

PrfA box

Code Strain Source Serotype hly actA prfA hly - mpl Internalin A structure PrfA 3d

LMO

175 monocytogenes L.

Isolated from Patient Blood

culture 4b *

+ - + + + -

+

LMO

179 monocytogenes L.

Isolated from Patient

Cerebrospinal fluid 4b *

+ - + + + -

+

LMO

(10)

LMO 34 monocytogenes L. Isolated from cheese 4b *

+ + + + + +

+

LMO 69 monocytogenes L. Isolated from milk 4b *

+ - - - + -

-

LMO

206 monocytogenes L. Isolated from cow´s liver 4b *

+

- + - + - +

LMO

195 monocytogenes L. Isolated from stool sample 1/2a

+ + - - + -

-

LMO 86 monocytogenes L. Isolated from siphon floor 4a

+ + + + + -

+

LMO

418 monocytogenes L. Isolated from cheese Unknown

+

+

+ - + - -

ATCC

7644 monocytogenes L. Reference ATCC Strain 7644 1/2c

+

+

+ - + + +

ATCC

19119 L. ivanovii Reference ATCC Strain 19119   

-

- - - -

ATCC

73016 L. innocua Reference ATCC Strain 73016   

-

- - - -

Table 4: Pathogenicity and PCR profiles of L.monocytogenes isolates. Virulence serotypes are indicated with an asterisk (*).

The actA gene amplified only in 60% of L. monocytogenes strains analyzed (Table 4). The lack of actA might be expected to affect virulence, since it is involved in cell to cell motility by polymerization of actin filaments of host cell. This is in agreement with the results obtained for the in vivo virulence assay (see below), as occurred with strains Lmo 135, Lmo 34 and Lmo 86 (Table 4). However, it is outstanding that strains Lmo175 and Lmo 179, which were isolated from patient blood culture and cerebrospinal fluid, respectively, had no amplification of actA

yet caused infection in patients. This may be due to divergence in the actA

sequence, such as reported insertion-deletions between nucleotides 852 to 1019 that result inalteration of a proline-rich repeat region[37, 52]. Variations in actA

sequence as previously mentioned could affect primer annealing and therefore

actA amplification.

These results also suggest the possible existence of mechanisms for L. monocytogenes spreading other than the requirement of a fully functional ActA protein. Indeed, a serovar 4b strain F2365 from a 1985 Jalisco cheese outbreak in California[53] has been shown to contain a 105-bp deletion in its actA gene, which did not seem to stop the bacterium causing listeriosis in humans[52, 54].

However, cases where actA deletion inhibits the expression of virulence have been also reported and it has been shown that a possible non-functional ActA may limit the ability to invade host cells [52]. This could be the case for Lmo 206, which did not amplify the actA gene and showed no invasion in the in vivo

(11)

virulence assays on HT 29 cells (see below), perhaps due to a defect in cell to cell spreading [52, 55].

Previous studies have identified the existence of polymorphisms in actA, with a rate between 14.3% [37]and 19.8% [54, 56, 57]. Therefore, it was also expected that some of the Colombian isolates analyzed here might show variation in this gene, as made evident by amplification via PCR. High variation in the actA gene sequence, such as insertion-deletions, can influence virulence in L. monocytogenes strains and is consistent with the observed results in this study where from three clinical isolates two actA negative strains were able to invade HT29 cells and one was not.

Most of the L. monocytogenes strains (80%) amplified the prfA gene, which is consistent with the importance of this gene as a central regulator of gene expression and with virulence. Strains Lmo 195 and Lmo69, which did not amplify

prfA, were not able to invade HT29 cell line (see below).

The Intergenic regions that had shown some variability in terms of prfA boxes by analyzing sequenced strains, hly – mpl and lmo02648 –inlA, were PCR-amplified and sequenced. The Intergenic region hly – mpl was amplified in 100% of the L. monocytogenes strains. Identity percentages of this intergenic region ranged between 91 and 98%, evidencing a low diversity among the isolates. Identical

prfA boxes were found in all evaluated strains, therefore positive regulation by PrfA is expected to occur in these isolates.

Nevertheless, the intergenic region upstream of inlA (lmo02648 –inlA) was amplified only in the reference strain L. monocytogenes ATCC 7644 and in the Lmo 34 isolate, with prfA boxes found in both cases. Although the lack of amplification is intriguing in the remaining L. monocytogenes strains, it is not possible to conclude that there are no prfA boxes. Further efforts should be made in order to improve primer design that allows the amplification of the remaining L. monocytogenes strain tested.

The overall assessment of virulence gene determinants in Colombian indicates that there is some variability present. This, however, might be further analyzed to confirm the results observed using strategies different from PCR as In Situ Hybridization. Amplification of virulence genes was independent of tested strain serotype. For example, all virulence genes were amplified in strain Lmo 86 with 4a non-virulent serotype, while only the hly gene was amplified in strain Lmo 69 with 4b virulent serotype. Even most of listeriosis outbreaks strains are classified as 4b, 1/2b and 1/2a serotypes, not all of them would necessarily be virulent because there is variability in the presence of genes associated with virulence. It is

(12)

necessary to improve virulence characterization techniques in order to find more precise determinants than serotypification.

PrfA modeling and Analysis

Given that we obtained prfA sequences for 8 of the Colombian strains analyzed, we decided to do a three dimensional structural analysis to see if there was variability in PrfA structure from Colombian isolates. Therefore three dimensional modeling was accomplished. Structure superposition was perfomed using each one of the models generated from the 8 translated prfA-amplified sequences and PrfAwt sequences downloaded from PDB (Protein Data Bank). Superposition showed 7 PrfA optimal structures and one (Lmo 418) nonfunctional PrfA structure (Figure 2). The white structure corresponds to PrfAwt and the colored structures correspond to the PrfA modeled proteins from the Colombian isolates.

Figure 2:Three-dimensional PrfA structure from L. monocytogenes isolates.

As can be seen, the PrfA three-dimensional structure was the same in all analyzed strains and corresponded well with the PrfAwt structure, except for Lmo 418. Based on this modeling, it appears that the Lmo 418 PrfA protein had no functional three-dimensional structure (Figure 2). As previously mentioned, the

prfA gene was not amplified from Lmo 69 and Lmo 195, and therefore no modeling was performed for these strains.

Two α-helices, αE (residues 169–174) and αF (184–195), and the connecting loop, αE - αF (174–184), constitute the helix–turn–helix (HTH) motif observed in many prokaryotic DNA-binding proteins, including DNA binding residues (183-195)[30]. All of these residues were conserved in the 7 PrfA modeled structures, indicating that the PrfA of Colombian strains has the potential to positively regulate

(13)

virulence genes. This is expected since these were isolated from clinical cases (Lmo 179 and 175) and is consistent with the observed virulence phenotype in vivo cell assays (see below) (Lmo 175, Lmo 179, Lmo 86, Lmo 34 and Lmo 135).

The surface of PrfAwt also reveals a distinct tunnel between the N-terminal β-barrel and the C-terminal DNA-binding domain of each monomer. This tunnel supports the hypothesis that PrfA like Crp may need an effector to induce activation[30]. The tunnel was observed in PrfA modeled proteins, except for Lmo 418 where the PrfA structure is incomplete (Figure 3).

Figure 3: Surface display and PrfA tunnel of L. monocytogenes isolates. The black circle indicates the location of PrfA tunnel.

PrfA has a direct relation with virulence. Strains Lmo 69, Lmo 195 and Lmo 418 with no amplification and no functional structure, respectively, were not able to invade HT-29 cells. This result confirms the central role of PrfA in expression of virulence genes during the intracellular cycle of L. monocytogenes. Since the actA gene was not amplified in Lmo 69 strain, it is difficult to establish if its non-virulent phenotype was due to the absence of prfA, actA or both of them.

In vivo virulence assay.

Virulence was assessed by analyzing the ability of L. monocytogenes strains to adhere to and invade HT29 cells. The first step in the infectious process is adhesion of the bacteria to the enterocytes[47], which is measured by washing cells and then freeing extracellularly adhered bacteria with Triton-X 0.1% and plating on Palcam Agar base. High adherence was obtained in all tested strains, including non-pathogenic controls L.

(14)

ivanovii ATCC 19119 and L. innocua ATCC 73016, as can be seen at the 1hr time point (Figure 4). However, differences were observed in terms of the ability to invade HT-29 cells. This is seen at later time points (4 and 6.5 h) by determining bacterial counts after eliminating extracellular bacteria with incubation of HT29 cells with DMEM supplemented with 100 ug/ml of gentamycin and then lysing host cells freeing internalized bacteria.

5 of 10 (50%) of the L. monocytogenes strains tested demonstrated the ability to invade HT-29 cells (Figure 4). This assay also made evident the lack of correlation between virulence and serotype since both virulent and non-virulent serotypes were able to invade and proliferate in the cell line.

Figure 4: L. monocytogenes virulence assay on HT29 cell line.

Of the three clinical isolates, two were able to invade host cells (Lmo 175 and Lmo 179 strains) and one was not (Lmo 195). As previously mentioned, Lmo 175 and Lmo 179 strains failed to amplify the actA gene. Despite this, their capacity to invade host cells was expected taking into account that they were isolated from blood samples and cerebrospinal fluid, indicating listeriosis, and therefore virulence, in humans. On the other hand, strain Lmo 195 was isolated from a patient stool sample and had no amplification of prfA. Therefore, it was not as surprising that it showed no virulence in the HT-29 cell line

(15)

assay, consistent with its presence in stool sample, a site not necessarily associated with virulence.

In general, the presence of all virulence genes tested (hly, actA and prfA) was associated with virulence in HT-29 cells, as observed in strains Lmo 135, 34 and 86. However, this was not always the case. For example, all three genes amplified in the L. monocytogenes ATCC 7644 strain, which nonetheless showed no virulence in our in vivo

assay. This is concordant with a previous study in which no virulence for this reference strain was observed either[48]. This reference strain 7644 has serotype 1/2c, which can be related to its non-virulence in cell lines and may explain these results[18]. Finally, strain Lmo 418 was not able to invade host cells, in agreement with the possible non-functional structure of PrfA obtained here by modeling (Figure 2), as mentioned previously.

The L. monocytogenes invasion of HT- 29 cells used in this study is dependent on both InlA and InlB proteins. In this order of ideas, it would be necessary to perform amplification of inlA and inlB genes to establish if the non-virulence observed in this study for strains Lmo 69, Lmo 206, Lmo 195 and L. monocytogenes ATCC 7644 could also be affected by the presence/absence of internalins[58].

The lack of correlation found in this study between the serotype of L. monocytogenes

strains and in vivo virulence or with the presence of virulence genes tested is concordant with previous reports where relative virulence in mice ranged between 37 and 93% using L. monocytogenes strains with virulent serotypes (4b, 1/2b and 1/2a). Further efforts should be made in order to establish if other virulence genes can be more reliable to assess the relationship between virulence and L. monocytogenes strain serotype.

In conclusion, the presence of virulence genes and of prfA boxes, as well as the predicted functionality of PrfA found in this study, as assessed by three-dimensional modeling, highlight the potential virulence of L. monocytogenes strains circulating in Colombia. Also, the finding of L. monocytogenes isolates from food, with one or all virulence genes leads to the conclusion that there is a high variability in circulating strains of this pathogen. It is important to continue working on this study with a representative number of strains isolated in Colombia, in order to determine how many of them can be truly pathogenic and therefore promote the legislation and mandatory notification of this pathogen in our country. This will allow to prevent and control possible outbreaks caused by this pathogen.

Moreover, the presence of the PrfA tunnel in most of the tested strains encourages the search of the cofactor used by L. monocytogenes, in order to develop control and inhibition techniques of the expression of the virulence genes of this pathogen in Colombia and worldwide.

(16)

Acknowledgments

We thank the Laboratory of Human Genetics from Universidad de los Andes for providing their facilities and give us support and Pablo Moreno of the National Cancer Institute by donating cell line HT-29. We thank the Investigations Committee (Sciences Faculty, Universidad de los Andes, Bogotá, Colombia) for their financial support.

References

1. Burbano E, Sierra S, Torres K, et al. (2006) Rapid DNA extraction adn PCR validation for direct detection of Listeria monocytogenes in raw milk. Revista MVZ Córdoba 11:715–724.

2. Ruiz-bolivar Z, Poutou-piñales RA, Carrascal-camacho AK Resistencia antimicrobiana y a desinfectantes de Listeria spp .

3. Vanegas MC, Vásquez E, Martinez AJ, Rueda AM (2009) Detection of Listeria monocytogenes in raw whole milk for human consumption in Colombia by real-time PCR. Food Control 20:430–432. doi: 10.1016/j.foodcont.2008.07.007

4. Carrascal-camacho AK, Castaño Sepúlveda MV, Cortes Muñoz MS, et al. (2011) Evaluación de Riesgo de Listeria monocytogenes en queso fresco. 1 – 67.

5. Ruiz-Bolivar Z, Neuque-Rico MC, Poutou-Piñales R a, et al. (2011) Antimicrobial susceptibility of Listeria monocytogenes food isolates from different cities in

Colombia. Foodborne pathogens and disease 8:913–9. doi: 10.1089/fpd.2010.0813

6. Camacho C, Contreras A, Torres S (2007) Incidencia de Listeria monocytogenes en leche de vaca expendida en el municipio de Pamplona , Colombia. Revista de la Facultad de Ciencias Básicas de la Universidad de Pamplona 5:49–57.

7. Belalcazar ME, García OT (2005) Listeria monocytogenes y Listeriosis animal. Revista UDCA 8:3 – 16.

8. Nataly S, Mercado M, Carrascal-camacho AK (2009) Efecto de tiempo y temperatura de cocción en chorizo inoculados artificialmente con Listeria monocytogenes. Universitas Scientiarum 14:198 – 205.

9. Adzitey F, Huda N (2010) Listeria monocytogenes in foods : Incidences and possible control measures. African Journal of Microbiology Research 4:2848–2855.

10. Gallegos J, Arrieta G, Salim M, et al. (2007) Frecuencia de Listeria spp ., en quesos colombianos costeños. Revista MVZ Córdoba 12:996–1012.

(17)

11. Torres K, Poutou R, Carrascal A, et al. (2004) PCR Validation for detection of Listeria monocytogenes in raw beef and chickens. Revista MVZ Córdoba 9:414–427.

12. Rahimi E, Ameri M, Momtaz H (2010) Prevalence and antimicrobial resistance of Listeria species isolated from milk and dairy products in Iran. Food Control 21:1448– 1452. doi: 10.1016/j.foodcont.2010.03.014

13. Albarracín Y, Poutou R, Carrascal-Camacho AK (2008) Listeria spp ., y L .

monocytogenes en leche cruda de cabra. Revista MVZ Córdoba 13:1326–1332.

14. Gandhi M, Chikindas ML (2007) Listeria: A foodborne pathogen that knows how to survive. International journal of food microbiology 113:1–15. doi:

10.1016/j.ijfoodmicro.2006.07.008

15. Report S, Efsa OF, Food E, et al. (2012) European Union Summary report Trends and Sources of Zoonoses , Zoonotic Agents and Food-borne Outbreaks in 2010. EFSA Journal 10:1–442. doi: 10.2903/j.efsa.2012.2597.

16. Scallan E, Mahon BE (2012) Foodborne Diseases Active Surveillance Network (FoodNet) in 2012: a foundation for food safety in the United States. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America 54 Suppl 5:S381–4. doi: 10.1093/cid/cis257

17. Muñoz AI (2012) Distribución de serotipos de Listeria monocytogenes aislados de alimentos, Colombia, 2000-2009. Biomedica 32:2000–2009.

18. Orsi RH, Den Bakker HC, Wiedmann M (2011) Listeria monocytogenes lineages:

Genomics, evolution, ecology, and phenotypic characteristics. International journal of medical microbiology : IJMM 301:79–96. doi: 10.1016/j.ijmm.2010.05.002

19. Gandhi M, Chikindas ML (2007) Listeria: A foodborne pathogen that knows how to survive. International journal of food microbiology 113:1–15. doi:

10.1016/j.ijfoodmicro.2006.07.008

20. Freitag NE, Port GC, Miner MD (2010) Listeria monocytogenes — from saprophyte to intracellular pathogen. Nat Rev Microbiol 7:1–15. doi: 10.1038/nrmicro2171.Listeria

21. Kreft J, Vázquez-Boland J a (2001) Regulation of virulence genes in Listeria.

International journal of medical microbiology : IJMM 291:145–57. doi: 10.1078/1438-4221-00111

22. Scortti M, Monzó HJ, Lacharme-Lora L, et al. (2007) The PrfA virulence regulon.

Microbes and infection / Institut Pasteur 9:1196–207. doi: 10.1016/j.micinf.2007.05.007

23. De las Heras A, Cain RJ, Bielecka MK, Vázquez-Boland J a (2011) Regulation of Listeria virulence: PrfA master and commander. Current opinion in microbiology 14:118–27. doi: 10.1016/j.mib.2011.01.005

(18)

24. Johnson J, Jinneman K, Stelma G, et al. (2004) Natural Atypical Listeria innocua Strains with Listeria monocytogenes Pathogenicity Island 1 Genes Natural Atypical Listeria innocua Strains with Listeria monocytogenes Pathogenicity Island 1 Genes. APPLIED AND ENVIRONMENTAL MICROBIOLOGY 4256–4266. doi:

10.1128/AEM.70.7.4256

25. Greene SL (2003) Negative regulation of PrfA, the key activator of Listeria monocytogenes virulence gene expression, is dispensable for bacterial pathogenesis. Microbiology 149:111–120. doi: 10.1099/mic.0.25692-0

26. De las Heras A, Cain RJ, Bielecka MK, Vázquez-Boland J a (2011) Regulation of Listeria virulence: PrfA master and commander. Current opinion in microbiology 14:118–27. doi: 10.1016/j.mib.2011.01.005

27. Eiting M, Hagelüken G, Schubert W-D, Heinz DW (2005) The mutation G145S in PrfA, a key virulence regulator of Listeria monocytogenes, increases DNA-binding affinity by stabilizing the HTH motif. Molecular microbiology 56:433–46. doi: 10.1111/j.1365-2958.2005.04561.x

28. Goebel W, Gonza B, Domi G, Patoge G De (2001) Listeria Pathogenesis and

Molecular Virulence Determinants. CLINICAL MICROBIOLOGY REVIEWS 14:584–640. doi: 10.1128/CMR.14.3.584

29. Nadon CA, Bowen BM, Wiedmann M, Boor KJ (2002) Sigma B Contributes to Mediated Virulence in Listeria monocytogenes Sigma B Contributes to PrfA-Mediated Virulence in Listeria monocytogenes. doi: 10.1128/IAI.70.7.3948

30. Eiting M, Hagelüken G, Schubert W-D, Heinz DW (2005) The mutation G145S in PrfA, a key virulence regulator of Listeria monocytogenes, increases DNA-binding affinity by stabilizing the HTH motif. Molecular microbiology 56:433–46. doi: 10.1111/j.1365-2958.2005.04561.x

31. Camacho C, Coulouris G, Avagyan V, et al. (2009) BLAST+: architecture and applications. BMC bioinformatics 10:421. doi: 10.1186/1471-2105-10-421

32. Morgulis A, Coulouris G, Raytselis Y, et al. (2008) Database indexing for production MegaBLAST searches. Bioinformatics (Oxford, England) 24:1757–64. doi:

10.1093/bioinformatics/btn322

33. Néron B, Ménager H, Maufrais C, et al. (2009) Mobyle: a new full web bioinformatics framework. Bioinformatics (Oxford, England) 25:3005–11. doi:

10.1093/bioinformatics/btp493

34. Notredame C, Higgins DG, Heringa J (2000) T-Coffee: A novel method for fast and accurate multiple sequence alignment. Journal of molecular biology 302:205–17. doi: 10.1006/jmbi.2000.4042

(19)

35. Baumgärtner M, Kärst U, Gerstel B, et al. (2007) Inactivation of Lgt allows systematic characterization of lipoproteins from Listeria monocytogenes. Journal of

bacteriology 189:313–24. doi: 10.1128/JB.00976-06

36. Lampel KA, Orlandi PA, Kornegay L (2000) Improved Template Preparation for PCR-Based Assays for Detection of Food-Borne Bacterial Pathogens Improved Template Preparation for PCR-Based Assays for Detection of Food-Borne Bacterial Pathogens. doi: 10.1128/AEM.66.10.4539-4542.2000.Updated

37. Cai S, Kabuki DY, Kuaye AY, et al. (2002) Rational Design of DNA Sequence-Based Strategies for Subtyping Listeria monocytogenes. Journal of Clinical Microbiology 40:3319–3325. doi: 10.1128/JCM.40.9.3319

38. Altschul SF, Madden TL, Schäffer AA, et al. (1997) Gapped BLAST and PSI-BLAST : a new generation of protein database search programs. Nucleic acids research 25:3389–3402.

39. Lovell SC, Davis IW, Iii WBA, et al. (2003) Structure Validation by C Geometry. Proteins: Structure, Function and Genetics 450:437–450.

40. Duhovny D, Nussinov R, Wolfson HJ (2002) Efficient Unbound Docking of Rigid Molecules. Proceedings of the 2’nd Workshop on Algorithms in Bioinformatics (WABI) 185–200.

41. Schneidman-Duhovny D, Inbar Y, Nussinov R, Wolfson HJ (2005) PatchDock and SymmDock: servers for rigid and symmetric docking. Nucleic acids research 33:W363–7. doi: 10.1093/nar/gki481

42. Johansson MU, Zoete V, Michielin O, Guex N (2012) Defining and searching for structural motifs using DeepView / Swiss-PdbViewer Defining and searching for structural motifs using DeepView / Swiss-PdbViewer. BMC Bioinformatics 173:1 – 10.

43. Guex N, Peitsch MC, Schwede T (2009) Automated comparative protein structure modeling with SWISS-MODEL and Swiss-PdbViewer: a historical perspective. Electrophoresis 30 Suppl 1:S162–73. doi: 10.1002/elps.200900140

44. Pettersen EF, Goddard TD, Huang CC, et al. (2004) UCSF Chimera--a visualization system for exploratory research and analysis. Journal of computational chemistry 25:1605–12. doi: 10.1002/jcc.20084

45. Yamada F, Ueda F, Ochiai Y, et al. (2006) Invasion assay of Listeria monocytogenes using Vero and Caco-2 cells. Journal of microbiological methods 66:96–103. doi: 10.1016/j.mimet.2005.10.017

46. Kushwaha K, Muriana PM (2010) Comparison of invasiveness among surface-adherent variants of Listeria monocytogenes in Caco-2 cell culture assays.

(20)

International journal of food microbiology 138:166–71. doi: 10.1016/j.ijfoodmicro.2009.11.024

47. Jensen A, Thomsen LE, Jørgensen RL, et al. (2008) Processing plant persistent strains of Listeria monocytogenes appear to have a lower virulence potential than clinical strains in selected virulence models. International journal of food microbiology 123:254–61. doi: 10.1016/j.ijfoodmicro.2008.02.016

48. Jaradat ZW, Bhunia AK (2003) Adhesion , Invasion , and Translocation Characteristics of Listeria monocytogenes Serotypes in Caco-2 Cell and Mouse Models. APPLIED AND ENVIRONMENTAL MICROBIOLOGY 69:3640–3645. doi: 10.1128/AEM.69.6.3640

49. Hain T, Steinweg C, Chakraborty T (2006) Comparative and functional genomics of Listeria spp. Journal of biotechnology 126:37–51. doi: 10.1016/j.jbiotec.2006.03.047

50. Goebel W, Gonza B, Domi G, Patoge G De (2001) Listeria Pathogenesis and Molecular Virulence Determinants. 14:584–640. doi: 10.1128/CMR.14.3.584

51. Chaturongakul S, Raengpradub S, Wiedmann M, Boor KJ (2008) Modulation of stress and virulence in Listeria monocytogenes. Trends in microbiology 16:388–96. doi: 10.1016/j.tim.2008.05.006

52. Chen J, Luo X, Jiang L, et al. (2009) Molecular characteristics and virulence potential of Listeria monocytogenes isolates from Chinese food systems. Food microbiology 26:103–11. doi: 10.1016/j.fm.2008.08.003

53. Nelson KE, Fouts DE, Mongodin EF, et al. (2004) Whole genome comparisons of serotype 4b and 1/2a strains of the food-borne pathogen Listeria monocytogenes reveal new insights into the core genome components of this species. Nucleic acids research 32:2386–95. doi: 10.1093/nar/gkh562

54. Conter M, Vergara A, Di Ciccio P, et al. (2010) Polymorphism of actA gene is not related to in vitro virulence of Listeria monocytogenes. International journal of food microbiology 137:100–5. doi: 10.1016/j.ijfoodmicro.2009.10.019

55. Ermolaeva S, Karpova T, Novella S, et al. (2003) A simple method for the

differentiation of Listeria monocytogenes based on induction of lecithinase activity by charcoal. International journal of food microbiology 82:87–94.

56. Medrano MV, Vanegas MC, Restrepo S (2006) Listeria monocytogenes en Colombia: Caracterización molecular por RAPD y Secuenciación del gen actA. 1 – 95.

57. Jiang L, Chen J, Xu J, et al. (2008) Virulence characterization and genotypic

analyses of Listeria monocytogenes isolates from food and processing environments in eastern China. International journal of food microbiology 121:53–9. doi:

(21)

58. Roche SM, Kerouanton a, Minet J, et al. (2009) Prevalence of low-virulence Listeria monocytogenes strains from different foods and environments. International journal of food microbiology 130:151–5. doi: 10.1016/j.ijfoodmicro.2009.01.012

Referencias

Documento similar

Functional analysis of intergenic regulatory regions of genes encoding surface adhesins in Staphylococcus aureus isolates from periprosthetic joint infections..

157 4.7.- Efecto de los electrones acelerados sobre la expresión de los genes de virulencia (plcA, hly e iap) y de respuesta al estrés (sigB) de Listeria monocytogenes en

In the present study, we investigated the presence of four Vibrio cholerae virulence genes (ctxA, VPI, Zot and ace) in 36 Vibrio alginolyticus isolates obtained from different

A pangenome analysis identi fied the presence or absence of gene clusters to be serotype speci fic, as well as virulence profile analyses targeting the apx operons..

Virulence genes and phenotypic features of the Klebsiella isolates analysed in this study: The presence of genes involved in virulence, i.e., magA and rmpA

coli isolates; Table S3: Antibiotic Resistance Genes (ARGs) and Virulence Factors (VFs) described by Abricate and CARD database; Table S4: IncF-family plasmids harboring

The Dwellers in the Garden of Allah 109... The Dwellers in the Garden of Allah

SmaI macrorestriction patterns in combination with analyses of the presence and expression of the described virulence-associated genes by PCR revealed a high level of similarity