Recombinant IgE-Reactive Functional Can f 5 Devoid of Cross-Reactive Carbohydrate Determinants

Abstract

Dogs are one of the most important allergen sources responsible for respiratory and skin allergy [1]. As early as 1970, dog epithelium was identified as a major allergen source responsible for allergic asthma [2]. Today, eight dog allergens have been described in the WHO-IUIS Allergen Nomenclature database (https://allergen.org/ accessed August 21, 2024) among which Can f 5 is highly important because it is a marker allergen specific for genuine dog allergy and especially for allergy to male dogs. According to studies carried out in different populations, between 15% and 20% of all IgE-sensitized subjects show IgE reactivity to Can f 5 [3-5]. Can f 5 is a 28-kDa prostatic kallikrein found in canine urine and fur. It is mainly produced in the prostate and its secretion is regulated by androgens [6]. Accordingly, expression of Can f 5 is observed only in male dogs. The nucleotide sequence coding for Can f 5 was first revealed in 1988 describing an androgen-dependent arginine esterase in the canine prostate which contains one potential N-glycosylation site [7]. So far, IgE-reactive recombinant Can f 5 allergen has been obtained only by expression in eukaryotic cells, in particular in yeast, which is known to add carbohydrate determinants which may be considered as cross-reactive carbohydrate determinants (CCDs) to proteins containing N-glycosylation sites [6]. IgE reactivity to CCDs often has no clinical relevance and CCD-specific cross-reactive IgE may react with CCDs present in a variety of unrelated allergen sources so that it may obscure the genuinely sensitizing allergen source [6, 8].

To the best of our knowledge, no IgE-reactive and functional, allergenic Can f 5 devoid of CCDs has been obtained by expression in Escherichia coli so far. The goal of our study was to obtain such non-glycosylated IgE-reactive Can f 5 exhibiting allergenic activity by expression in E. coli. To this end, several expression constructs were prepared (Figure 1, Table S1). They comprise a recombinant construct containing the N-terminal leader (pre-) and pro-peptide with a C-terminal hexahistidine tag (rPrePro Can f 5-His), an identical construct with one single N-terminal aspartic amino acid (rAsp-PrePro Can f 5-His), a construct comprising Can f 5 with a N-terminal hexahistidine tag and the prepro-peptide (rHis-PrePro Can f 5), and finally a Can f 5 construct containing the N-terminal pro-peptide and a C-terminal hexahistidine tag (rPro Can f 5-His) (Figure 1). The characterization of these constructs and their expression are reported in the Supporting Information. Figure S1 exemplifies for three different clones that rPrePro Can f 5-His could not be expressed in E. coli, whereas rAsp-PrePro Can f 5-His (Figure S2), rHis-PrePro Can f 5 (Figure S3), and rPro Can f 5-His (Figure S4) were expressed upon induction of protein expression with IPTG. Figure 2a shows that rAsp-PrePro Can f 5-His shows comparable IgE reactivity as rPro Can f 5-His, whereas no IgE-reactive rPrePro Can f 5-His could be expressed. This result was surprising because rAsp-PrePro Can f 5-His differed from rPrePro Can f 5-His only by an additional aspartic acid at the N-terminus (Figure 1) which may indicate that the addition of hydrophilic amino acids in front of N-terminal hydrophobic sequences in a protein (e.g., Pre- and Pro-sequences in Can f 5) may facilitate their expression in E. coli. This assumption is supported by the finding that the addition of a hydrophilic hexahistidine tag to the N-terminus of PrePro Can f 5 also allowed the expression of the protein (Figures S3) and the rHis-PrePro Can f 5 reacted specifically with IgE antibodies from Can f 5 sensitized patients (data not shown). Figure 2a shows that rAsp-PrePro Can f 5-His such as the positive control, rPro Can f 5-His, but not rPrePro Can f 5-His exhibit specific IgE reactivity with sera from each of the Can f 5-sensitized patients tested.

Escherichia coli-expressed rPro Can f 5-His could be produced in large quantities and assumed secondary structure according to circular dichroism (Figure S5) and therefore seemed to be a suitable candidate molecule for IgE testing. It was purified and showed specific IgE reactivity when tested with sera from dog-sensitized subjects that were positive when tested by ImmunoCAP ISAC (Thermo Fisher, Uppsala, Sweden) (data not shown, Table S2). Importantly, rPro Can f 5-His induced specific degranulation of basophils loaded with serum IgE from Can f 5-sensitized patients, but not when sera from nonallergic subjects were used (Figure S6 and Table S2). In a second set of basophil activation experiments, we show that rPro Can f 5-His and yeast-expressed Can f 5 have comparable allergenic activity in Can f 5-sensitized patients (Figure 2b, panel A). However, basophil activation was observed only in patients sensitized to rPro Can f 5-His, but not in patients who, according to IgE inhibition experiments performed with yeast-derived CCD, reacted only with CCDs (Tables S2 and S3, Figure 2b, panel B). Thus, basophil experiments indicate that E. coli- and yeast-expressed Can f 5 exhibit comparable allergenic activity in basophil activation experiments and that only IgE-reactive CCDs present on yeast-expressed Can f 5 lack allergenic activity (Figure 2b).

The fact that Can f 5-sensitized subjects can be found who show not only IgE reactivity to the Can f 5 protein but also to CCDs is shown by IgE inhibition experiments performed by ELISA (Table S3). IgE binding of serum from patient 1 to yeast-expressed rCan f 5 was inhibited by 36% by pre-incubation with yeast extract, whereas this was not the case for other Can f 5-sensitized subjects (Table S3). Except for patient 1, IgE levels of E. coli-expressed rPro Can f 5 were comparable to those of yeast-expressed Can f 5 after preincubation with yeast extract (Table S2). Among 110 subjects exhibiting IgE reactivity to CCDs without symptoms of dog allergy, four patients were identified who, according to IgE inhibition experiments, reacted only with yeast CCDs but not with the proteinaceous part of Can f 5 (Table S3, patients 12, 16, 17, and 18). This finding is important regarding measuring specific IgE reactivity to Can f 5 in serological allergy tests.

To the best of our knowledge, purified Can f 5 is available only in ImmunoCAP and ImmunoCAP ISAC commercial allergy tests, but we could not find information on what type of purified Can f 5 (e.g., purified natural, yeast-, or E. coli-expressed Can f 5) is used in these tests. Other multiplex assays such as ALEX (https://www.macroarraydx.com/de/produkte/alex) contain only dog urine but not the purified CCD-free Can f 5. Accordingly, there is a need for IgE tests containing Can f 5 devoid of IgE-reactive CCD epitopes.

Molecular diagnosis of dog allergy has important clinical implications because differences in clinical phenotypes have been reported for patients depending on molecular IgE sensitization profiles, which may affect allergen-specific forms of treatment such as allergen-specific immunotherapy (AIT) [9]. For example, patients with Can f 5 sensitization may tolerate female dogs better because they do not express the Can f 5 allergen [10]. Furthermore, molecular immunotherapy vaccines need to be designed to contain all relevant dog allergens, and it will therefore be necessary to identify the relevant dog allergens using functional recombinant molecules, as has been shown for cat allergens [11, 12].

In summary, we describe the expression of non-glycosylated IgE-reactive and allergenic Can f 5 in E. coli, which is suitable for measuring IgE specific for the proteinaceous moieties of Can f 5 and which avoids cross-reactivity with CCDs. Thus, E. coli-expressed Can f 5 allows identification of patients with a genuine sensitization to dog and in particular of those allergic to male dogs.

Evgenii Kozlov: designed and performed experiments, analyzed data, wrote the manuscript, and read the manuscript; Daria Trifonova, Alexandra Dubovets, Anastasia Usanova: Performed experiments, analyzed data, and read the manuscript; Daria Trifonova, Inna Tualeva, Pia Gattinger: Performed experiments, analyzed data, and read the manuscript; Rudolf Valenta, Neonila Gorokhovets, Alexander Karaulov: Analyzed data, wrote the manuscript, read the manuscript, and designed and supervised experiments. Daria Fomina, Wolfgang Hemmer: characterized patients, provided samples, analyzed data, and read the manuscript.

Anonymized sera from patients were analyzed with the permission of the ethics committee of the Medical University of Vienna (EK 1641/2014) after informed consent was obtained. Sera from Sechenov First Moscow State Medical University were obtained and used after approval by the local ethics committee (№ 04-22, dated February 16, 2022).

Rudolf Valenta has received research grants from HVD Life-Sciences, Vienna, Austria, and from WORG Pharmaceuticals, Hangzhou, China. He serves as a consultant for HVD and WORG Pharmaceuticals. The authors with a Russian affiliation declare that they have prepared the article in their “personal capacity” and/or that they are employed at an academic/research institution where research or education is the primary function of the entity. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results. The other authors have no conflicts of interest to declare.