why do dogs need hydrolyzed protein

Hydrolyzed diets may stimulate food-reactive lymphocytes indogs

Elimination diets often contain low molecular weight (MW) proteins and peptides resultingfrom enzymatic digestion (hydrolysis) of proteins from source materials [2, 10, 20]. Hydrolyzed protein diets are considered therapeutic for companionanimals since they may prevent allergic reactions due to food hypersensitivity [2]. In humans, IgE may recognize protein allergens with MWranging from 550 kDa, therefore, hydrolyzed proteins in elimination diets must have MW lowerthan 5 kDa to prevent allergic reactions [29]. A recentstudy in dogs demonstrated that a canine diet made from hydrolyzed poultry-feather mealcontaining 95% hydrolyzed proteins with MW 1 kDa did not induce clinical reactions in any ofthe dogs allergic to chicken. However, another diet containing 78% hydrolyzed chicken liverproteins with MW 1 kDa did induce allergic reactions in 40% of the dogs [2].

A clinical study of dogs diagnosed with food hypersensitivity after food restriction andprovocation found type IV hypersensitivity lymphocyte-mediated reactions in 82% of the dogs,however, type I hypersensitivity was rarely detected [8]. Another report evaluating clinical cases of dogs with chronic dermatitis andsuspected food hypersensitivity confirmed this trend [5]. In addition, a study conducted on dogs with gastrointestinal symptoms clinicallydiagnosed as inflammatory bowel disease, found peripheral lymphocytes reactive to at least onefood antigen in all the animals [12]. Flow cytometryanalysis revealed that the type IV lymphocytes associated with hypersensitivity wereCD4⁺ helper T-lymphocytes [5, 21]. Theoretically, these types of lymphocytes are reactiveto proteins or peptides that do not induce IgE-mediated reactions [10, 14]. Therefore, it is clear thatclinical use of hydrolyzed diets, as part of food elimination programs, may not be appropriatefor all dogs with food hypersensitivity.

Small peptides (T cell epitopes) recognized by helper T-lymphocytes are displayed by MHC(Major Histocompatibility Complex) class II molecules found on antigen-presenting cells, suchas macrophages, dendritic cells, and B-lymphocytes, which engulf antigens via phagocytosis. Inprevious studies, Wu et al., 2002, and Masuda et al. 2004,experimentally determined the amino acid (AA) sequences of common allergens by mapping withoverlapping peptides containing ca. 15 AAs covering all the protein allergensequences [17, 30]. The estimated MW of the 15 AA length peptides ranged from 13 kDa [6, 17, 23], depending on AA composition. In a 2007 study, Kabukiand Joh showed that a peptide MW range of 13 kDa was critical for induction of a T-lymphocytereaction in a human infant with casein-reactive T-lymphocytes. The study results demonstratedthat helper T-lymphocytes were stimulated by mildly hydrolyzed casein with MW of 2 kDa, butnothighly hydrolyzed casein with MW 1 kDa [10].Moreover, in another study, a specific T-lymphocyte helper clone could recognize a peptideconsisting of only 5 AA, with a MW <1 kDa [7]. Theseresults suggest that helper T-lymphocytes may recognize hydrolyzed proteins in a clinicalsetting, therefore, lower hydrolyzed-protein MW are desirable to prevent reactions.

Since T-cell epitopes are linear peptides, allergen AA sequences may be conserved in similarspecies, and therefore cause allergic reactions due to T-lymphocyte cross-reactivity. In arecent study, the major allergens from Japanese cedar and cypress Cry j 1 and Cha o 1,contained AA sequences with approximately 80% homology [26], and also contained identical T-cell epitope sequences recognized by humanT-lymphocytes reactive to both Cry j 1 and Cha o 1 [23]. The Cry j 1 and Cha o 1 T-cell epitopes were also identical to various treepollen and vegetable species epitopes [9, 31]. Alpha-parvalbumin, a major allergen from chicken meat,contains AA sequences that are highly conserved (greater than 80% homology) in poultry andlivestock animals, and in a previous study resulted in induction of cross-reactive human IgE[15]. In 2018, Bexley et al.reported that dog serum IgE cross-reactivity between chicken and fish was related to nineproteins common to both species [1]. Alpha-actin, ahighly conserved gene in vertebrates, was identified as one of the nine proteins [3]. Consequently, since T-lymphocytes may be widelycross-reactive to low MW allergens from taxonomically similar species, it is possible thatT-lymphocytes could react to proteins or peptides even after hydrolysis.

In this study, the MW and antigenicity of protein extracts from two types of hydrolyzeddiets, Aminopeptide Formula Dry (Royal Canin Japon) and Canine z/d Ultra Dry (Hills-Colgate(Japan) Ltd.), were examined to determine whether hydrolyzed proteins could stimulate helperT-lymphocytes in dogs with suspected food hypersensitivity.

MATERIALS AND METHODS

Protein extraction from hydrolyzed diets

Two types of commercially available hydrolyzed diets (D-1, Aminopeptide Formula Dry,Royal Canin Japon, known as Anallergenic, Royal Canin, Aimargues, France and D-2, HillsPrescription Diet, Canine z/d Ultra Dry, Hills-Colgate (Japan) Ltd.) were purchased. Whenthe study was initiated, three product lots of D-1, and two lots of D-2 were available.The protein extraction protocol was similar to the one previously reported [11]. Briefly, one gram of crushed kibbles from eachdiet was minced and suspended in 10 ml of physiological saline, andsubsequently ground with glass beads under shaking conditions at a speed of 6,000 rpm for1 min, in an ULTRA-TURRAX Tube Drive (IKA Werke GmbH & Co., KG., Staufen,Germany). This procedure was repeated three times with one-min intervals. The suspensionwas maintained at 4C for 30 min to remove sediments after homogenization, and thesupernatant was collected and centrifuged at 10,000 g for 10 min at 4C. The supernatantwas filtered and sterilized after centrifugation, with a 0.45 m poresize PES membrane, and protein concentrations were measured with ultraviolet (UV) lightabsorbance at a wavelength of 280 nm. Finally, the extracts were analyzed with sodiumdodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and size-exclusionchromatography, and tested for lymphocyte stimulation.

SDS-PAGE

The hydrolyzed diet and chicken meat extracts (Greer Laboratories, Inc., Lenoir, NC, USA)were separated by SDS-PAGE in a 520% gradient polyacrylamide gel (ATTO Corp., Tokyo,Japan) using a previously described method [11].The total protein amounts in D-1, D-2, and chicken meat extract were adjusted to 10g per lane. After electrophoresis, the gel was stained with Coomassiebrilliant blue, (CBB) (FUJIFILM Wako Pure Chemical Corp., Osaka, Japan), or silver stainreagent (Cosmo Bio Co., Ltd., Tokyo, Japan) for protein band visualization.

Size-exclusion chromatography

Size-exclusion chromatography of the extracts was performed using slight modifications ofa previously reported procedure [11]. A Superdex200 10/300 GL column (GE Healthcare Bio-Sciences AB, Uppsala, Sweden) was equilibratedwith phosphate buffered saline (PBS). Each sample was eluted isocratically at a flow rateof 0.5 ml/min in PBS, and protein elution was monitored with UVabsorbance at 280 nm. The column was calibrated with a Gel Filtration Low MW Kit (GEHealthcare Bio-Sciences AB), containing different MW protein standards: ferritin (MW 440kDa), aldolase (MW 158 kDa), conalbumin (MW 75 kDa), ovalbumin (MW 43 kDa), carbonicanhydrase (MW 29 kDa), ribonuclease A (MW 13.7 kDa), and aprotinin (MW 6.5 kDa).

Lymphocyte stimulation

As previously reported [5, 12, 25], lymphocyteproliferation of PBMCs isolated (by veterinarians in private animal hospitals) from dogswith suspected food hypersensitivity, was investigated in a routine assay with 18 types offood antigens in a commercial laboratory (Animal Allergy Clinical Laboratories, Inc.,Sagamihara, Japan). The food allergens for the standard lymphocyte assay were extractedfrom beef, pork, chicken, egg white, egg yolk, milk, wheat, soybean, corn, lamb, turkey,duck, salmon, cod fish, catfish, potato, and rice, and all were prepared and purchasedfrom Greer Laboratories Inc. as previously described [5, 12, 25]. However, capelin was prepared by homogenization of the dried fish body inPBS. Detailed clinical symptoms and dog history were unknown, however, dermatologicaland/or gastrointestinal symptoms were suspected. Additionally, no information on thestatus of allergen exposure and treatment history was available for the dogs providing theblood samples. The blood samples were kept at 4C and shipped to the laboratory byrefrigerated courier services within 13 days after collection from local veterinarians.Briefly, PBMCs were purified from whole blood with Ficoll-Hypaque (Lymphoprep,Axis-shield, Oslo, Norway) under centrifugation. The cell viability was determined byexclusion of trypan blue (Sigma-Aldrich, Inc., St. Louis, MO, USA). When more than 95% ofcells were viable, the PBMCs were cultured for seven days in RPMI-1640 medium (FUJIFILMWako Pure Chemical Corp.), containing 10% fetal bovine serum, 100g/ml streptomycin, and 100 U/mlpenicillin (Sigma-Aldrich), at 1 g/ml antigenconcentration for each allergen. Following the standard assay with 18 types of foodallergens, the remaining PBMCs were also tested in a similar manner with extracts fromD-1 and D-2. On Day 4 of the culture, recombinant human IL-2 (Pepro Tech Inc., Rocky Hill,NJ, USA) was added at a concentration of 100 U/ml. Subsequently, thecultured PBMCs were stained with antibodies, and analyzed via flow cytometry.

Flow cytometry

As previously described [5, 14], the antibodies used in the flow cytometry analysis includedanti-dog CD4 rat antibody labeled with Alexa Fluor 647 (Clone YKIX302.9) (Bio-RadLaboratories, Inc., Hercules, CA, USA), and anti-human CD25 antibody labeled withphycoerythrin (Clone ACT-1) (DAKO Cytomation, Glostrup, Denmark), as well as isotypecontrols such as Alexa Fluor 647-labeled purified rat IgG2a (Bio-Rad Laboratories), andPE-labeled mouse IgG1 (Bio-Rad Laboratories). Stained dead cells were excluded bypropidium iodide incorporation. Finally, CD4⁺CD25⁺ cells weremeasured in a lymphocyte gate with FACS Canto II and FACS Diva software (Becton Dickinson,San Jose, CA, USA). We determined the number of activated helper T-lymphocytes againsteach antigen by calculating the percentage of CD4⁺ CD25^lowlymphocytes in 2500 CD4⁺ lymphocytes [18, 19]. We defined activation as 0.4%CD4-positive T-lymphocytes, since at least ten positive cells (0.4% of 2,500 cells)confirmed the presence of a distinct cell population.

When considering clinical use of hydrolyzed diets in a food elimination program,induction of a lymphocyte response in dogs allergic to poultry-related antigens should bea major concern, since the diets examined in this study contained poultry-relatedproteins, such as feather meal or chicken meat. In this study, the positive lymphocyte(0.4%) response to poultry-related antigens was compared to the response induced withhydrolyzed proteins. When a positive lymphocyte response to two or more poultry-relatedantigens was detected in a sample, the maximum antigen value was designated asrepresentative of a lymphocyte response. Positive lymphocyte responses to poultry-relatedantigens and hydrolyzed diets were statistically analyzed with Spearmans rank correlationcoefficient. In addition, since positive lymphocyte responses might not always bedetected, qualitative comparisons were also conducted. The ² test was utilizedto analyze categorical data from samples with positive (0.4%) and negative (<0.4%)lymphocyte responses, and from poultry-related antigens and hydrolyzed diets. Statisticalanalyses were conducted with XLSTAT software (Addinsoft, Paris, France).

A value of 1.2% has been considered a threshold of clinical manifestation in dogs withfood hypersensitivity [5]. Therefore, we counted thenumber of samples with hydrolyzed diet response values of 1.2%, to calculate a lymphocyteresponse potentially associated with clinical manifestation.

RESULTS

Hydrolyzed diet protein molecular weights

SDS PAGE gels of the D-2 extract processed with CBB and silver stain revealed multiplebroad bands with MW from 3.524 kDa, and faint bands at approximately 24 kDa and 52 kDa.No bands were detected in the D-1 extract processed with CBB, and silver staining revealedonly one faint band with a MW slightly larger than 24 kDa (). Since the protein amounts in all lanes (D-1, D-2 and chicken meat) were adjustedto 10 g, the differences in visualized bands were due to proteincomponent differences in the samples.

SDS-PAGE of D-1 (Lane 1), D-2 (Lane 2), and chicken meat (Lane 3) extracts withCoomassie brilliant blue (CBB) and silver staining. The total protein amount appliedto each lane was 10 g. Note that there were no visualized bands inLane 1. SDS-PAGE, sodium dodecyl sulphate-polyacrylamide gel electrophoresis.

Size-exclusion chromatography revealed 68 protein waveform peaks with differentmolecular weights in both D-1 and D-2 ( and ). Based on the standard protein reagents, the two major peaks had an estimated MW>1 kDa, approximately 1.5 kDa and 3.5 kDa. A peak with a MW >440 kDa (the largeststandard protein) was detected in the D-2 extracts. All three D-1 product lots, and twoD-2 product lots displayed similar wave patterns (). Size-exclusion chromatography data obtained in a different laboratory (TorayResearch Center, Inc., Tokyo, Japan), revealed waveform peaks similar to our results (datanot shown).

Size-exclusion chromatography of D-1 extracts. The assay was repeated with threedifferent product lots. Similar wave patterns were detected in all product lots.

Size-exclusion chromatography of D-2 extracts. The assay was repeated with twodifferent product lots. Similar wave patterns were detected in all product lots.

Lymphocyte responses to hydrolyzed proteins and poultry-related antigens

Of 316 PBMC total samples examined in the study, positive lymphocyte responses (0.4%)against D-1 and D-2 extracts were detected in 91 (91/316, 28.8%), and 75 (75/316, 23.7%)samples, respectively ().

Table 1.

In each sample, the presence or absence of helper T-lymphocytes reactive tohydrolyzed foods (D-1 and D-2) and poultry proteins was compared. Samples werestratified at 0.4% of reactive helper T-lymphocytes, since it was considered in thisstudy that 0.4% was the minimum percentage to show reactive helperT-lymphocytes

We applied Spearmans rank correlation coefficient (due to lack of normality) todetermine if there was a correlation between positive lymphocyte responses in the D-1 orD-2 extracts, and the maximum response with poultry-related antigens. However, nocorrelations were observed, (D-1, rs=0.256; D-2, rs=0.239). Positive lymphocyte responses(0.4%) were observed with at least one poultry-related antigen, such as chicken, duck,turkey, or chicken egg, in 186 of 316 total samples (186/316, 58.9%). The D-1 extractselicited positive lymphocyte responses in 72 samples (72/186, 38.7%), and the D-2 extractselicited responses in 55 (55/186, 29.6%) samples (). D-1 and D-2 extracts elicited positive responses in 130 samples thatdid not exhibit positive lymphocyte responses to any poultry-related antigens; 20 samples(20/130, 15.4%) responded to D-1, and 19 samples (19/130, 14.6%) to D-2. Application ofthe Chi-squared test demonstrated that there was no association between the maximumpercentage of positive lymphocyte responses to poultry-related antigens, and the positiveresponses to the D-1 or D-2 extract.

Only seven (7/316, 2.2%) D-1 samples and six D-2 samples (6/316, 1.9%) elicitedlymphocyte reactions with values 1.2%.

DISCUSSION

Hydrolyzed diets are considered effective for food elimination programs in companionanimals with food hypersensitivity, since they contain proteins or peptides with molecularweights that are typically too low to stimulate allergic reactions. However, clinical casesof dogs with food hypersensitivity that did not respond positively to food eliminationprograms with hydrolyzed diets have been reported [2].In addition, a previous report revealed that several carbohydrate proteins with molecularweights ranging from 2167 kDa, and those recognized by serum IgE from individual cases ofdogs with suspected food hypersensitivity, were detected in the same commercial dietsexamined in this study [22]. In this study, SDS-PAGEgels of D-1 and D-2 extracts revealed bands with slightly higher MW than 24 kDa. Inaddition, a peak with an extremely high MW (>440 kDa) was detected only in D-2 extractswith size exclusion chromatography. These results suggest that type I hypersensitivity mayoccur in clinical cases where serum IgE is reactive to the proteins.

Previous reports showed that a type IV hypersensitivity allergic reaction mediated byT-lymphocytes recognizing small peptides not recognized by IgE, was observed in dogs withfood hypersensitivity [5, 8]. In this study, the lower MW proteins and peptides in the hydrolyzeddiet extracts were tested for induction of lymphocyte-mediated allergic reactions.Size-exclusion chromatography of D-1 and D-2 extracts revealed proteins and peptidesapproximately 1.5 kDa and 3.5 kDa MW ().Previous studies suggested that proteins or peptides of this size were large enough toinduce helper T-lymphocyte reactions, since synthetic peptides between ca.11.5 kDa MW were used in experiments to detect helper T-lymphocyte reactions to T-cellepitopes [13, 18]. In addition, flow cytometry results in this study demonstrated a lymphocyteresponse to low MW proteins and peptides in the hydrolyzed diets. These results confirm thateven low MW proteins and peptides may not be small enough to avoid a T-lymphocyte responsein dogs with food hypersensitivity. Consequently, although hydrolyzed diets have beenclinically effective in food elimination programs for many dogs with food hypersensitivityas previously reported [2], it is possible thatT-lymphocyte allergic reactions might be prompted by residual proteins in hydrolyzed diets,and therefore these diets should not be considered appropriate for all dogs with foodhypersensitivity.

Although potential induction of T-lymphocyte reactions against hydrolyzed protein diets hasbeen suspected, induction of clinical symptoms related to food hypersensitivity in dogs wasconsidered unlikely for three reasons. First, in this study, T-lymphocyte responses withvalues 1.2% were detected in only seven of 316 (2.2%), and six of 316 (1.9%) casesstimulated by D-1 and D-2, respectively. Since 1.2% has been considered the threshold forclinical manifestation of canine food hypersensitivity [5], hydrolyzed diets should be safe in most cases. Second, in dogs fed hydrolyzeddiets, gastrointestinal digestion should decrease the molecular size of proteins orpeptides, possibly preventing antigenicity against T-lymphocytes. For example, frequentgastrointestinal digestion of collagen in vitro results in small peptideswith MW <1 kDa [24], a molecular size consistentwith common T-cell epitopes [17, 23]. Therefore, any T-cell epitopes remaining inhydrolyzed protein diets should be further reduced in size during gastrointestinaldigestion. Third, since the protein sources in the hydrolyzed diets examined in this studywere poultry-feather meal (D-1), and chicken meat (D-2) [2], they may have included T-cell epitopes taxonomically similar to many poultryproteins. In general, T-cell epitopes are conserved among major allergens with taxonomicsimilarity [28], and cross-reactivity againstmultiple allergens can occur [9, 23, 31]. T-cell epitopes should beconserved in poultry-related allergens; this may explain the bird-egg syndrome phenomenonof human allergic reactions to some avian meats [13]and feathers [27]. However, in this study, we foundno significant correlation between positive lymphocyte responses against D-1 or D-2extracts, and poultry-related allergens, suggesting that cross-reactivity of T-lymphocytereactions may not be very common. Although further study on T-cell epitopes ofpoultry-related antigens in dogs is necessary, at present, hydrolyzed diets are consideredto have a low risk for T-lymphocyte cross-reactivity. Consequently, a food eliminationclinical program using hydrolyzed diets should be successful in most dogs with foodhypersensitivity.

We detected one faint but clear protein band slightly larger than 24 kDa MW in silverstained SDS-PAGE gels of D-1 and D-2 extracts. This was inconsistent with results from aprevious report where wide ranges of residual proteins from 2167 kDa MW were found in thesame hydrolyzed diets examined in this study [22].Possibly, this may be due to differences in protein extraction procedures. The previousreport described a homogenization procedure with a tissue homogenizer,versus the easy-to-use tabletop homogenizer utilized in the presentstudy. In addition, protein extraction efficiency may be increased when the startingmaterial is pulverized rather than homogenized [16],therefore an optimized method for protein extraction from hydrolyzed diets should bedeveloped.

In this study, protein concentrations were measured with UV light absorption at 280 nmbecause dye-staining methods, such as the Bradford method, were unsuccessful for the D-1extracts (data not shown). This result was consistent with the SDS-PAGE results since CBBstaining did not visualize any D-1 extract proteins in the lower MW ranges (3.524 kDa),whereas broad protein bands were visualized with both CBB and silver stain in the D-2extracts. It is likely that proteins were not stained in the D-1 extracts due to structuraldamage of AA. CBB reacts with many types of AA, including aromatic and basic, as well aswith the AA N-terminus [4], likely destroyed in theD-1 extract. In comparison, during silver staining, silver ions bind to AA sulfhydryl andhydroxyl groups, subsequently visualized with paraformaldehyde in a basic pH solution thatreduces silver ions to metallic silver [17]. Notably,the D-1 extract proteins were also not visualized with silver stain, suggesting that the AAsulfhydryl and/or hydroxyl groups might have been damaged in the extraction procedure. Inaddition, regardless of the negative staining results, the D-1 extract proteinconcentrations were easily measured with UV light, which is absorbed by AA benzene rings.These results suggest that AA conformation in the D-1 extracts might have been partiallydamaged, but not destroyed. However, since proteins in the D-2 extracts were detected afterboth CBB and silver staining; the protein extraction procedure could not have affectedprotein conformation in the D-1 extracts. Therefore, proteins in the D-1 source material,i.e. hydrolyzed feather meal, should be analyzed to determine whetherhydrolyzation could potentially cause changes in protein conformation.

Since the lymphocyte response is influenced by several factors in cell culture conditionssuch as reactive antigen or peptide concentrations, duration of antigen stimulation, cellconcentrations, and ratios of reactive lymphocytes and antigen-presenting cells,T-lymphocyte reactions would not necessarily always be detected in this study.False-negative results might occur depending on cell culture conditions; although alymphocyte response 1.2% was designated as the food hypersensitivity detection threshold indogs [5, 9].Therefore, the significance of this study was the in vitro detection oflymphocyte responses in dogs stimulated by hydrolyzed diet extracts, previously reported inhumans [10], but not in dogs. Lymphocyte responseswould be controllable in constant culture conditions, and comparable to lymphocyte responsesstimulated by different antigens in other studies. These results could be obtained in dogs,(as in studies of human helper T-lymphocyte proliferation), by determination of the specificco-culture ratio between antigen-presenting cells, and CD4 positive lymphocytes,autologously isolated from blood samples [1, 3].

The PBMCs utilized in this study were sent by veterinarians for a previously describedstandard lymphocyte proliferation assay [5, 12, 25], and afterthe assay was completed, the remaining cells were used to examine T-lymphocyte responses tothe hydrolyzed diets. Therefore, we did not have access to the history and clinical featuresof the dogs and blood samples. Ideally, before blood samples are submitted for lymphocyteproliferation assays, dogs with suspected allergies should be clinically diagnosed with foodhypersensitivity based on their responses to food elimination programs and provocation[8]. Typically, only a limited number of dogs may beaccepted for this type of analysis since the diagnostic procedure is complex, and may notalways be successfully completed. In this study, we examined 316 samples, and although thesample inclusion criteria were not strictly defined, the large sample size permitted acomprehensive analysis of T-lymphocyte responses to hydrolyzed diets in dogs with potentialfood hypersensitivity. However, further studies should be conducted to confirm T-lymphocytereactions to hydrolyzed diets in dogs with food hypersensitivity.

In summary, the current study demonstrated that hydrolyzed diets contain proteins andpeptides with sufficiently high molecular weights to stimulate T-lymphocytes. Therefore, itis important to avoid hydrolyzed diets in food elimination programs when treating dogspreviously diagnosed with lymphocytes reactive to poultry-related antigens.