The Paramyxoviridae family includes some of the most important and ubiquitous disease-causing viruses of infants and children, most of which cause significant infections of the respiratory tract. Evidence is accumulating in humans that genetic factors are involved in the severity of clinical presentation. As a first step toward the identification of the genes involved, this study was undertaken to establish whether laboratory mouse strains differ in susceptibility to Sendai virus, the murine counterpart of human type-1 parainfluenza virus which, historically, has been used extensively in studies that have defined the basic biological properties of paramyxoviruses in general. With this purpose in mind, double-chamber plethysmography data were collected daily for 7 days after inoculation of Sendai virus in six inbred strains of mice. In parallel, histological examinations and lung viral titration were carried out from day 5 to day 7 after inoculation. Pulmonary structure/function values closely reflected the success of viral replication in the lungs and revealed a pattern of continuous variation with resistant, intermediate, and susceptible strains. The results unambiguously suggest that BALB/c (resistant) and 129Sv (susceptible) strains should be used in crossing experiments aimed at identifying the genes involved in resistance to Paramyxoviridae by the positional cloning approach.
- genetic resistance
the paramyxoviridae family includes some of the great and ubiquitous disease-causing viruses of infants and children. Evidence is accumulating in humans that genetic factors are involved both in the control of infectious diseases and in the regulation of infection levels and clinical presentation, as demonstrated by racial differences in susceptibility to infection (1, 2, 4, 26, 38) and familial occurrence of infection in endemic areas (20, 32, 33). Thus identifying genes that control the lung response to paramyxoviruses is a crucial step in elucidating how they might affect the pathophysiological processes underlying the severity of the disease induced. Because susceptibility to infection is usually a complex genetic trait, the challenge is to dissect this into putatively monogenic phenotypes. This can be done most efficiently by using strains of laboratory mice. Here, the first step in isolating such genes is to identify strains that differ in their susceptibility to the pathogens targeted.
The present study aimed at implementing this strategy with Sendai virus (SeV), the murine counterpart of human type-1 parainfluenza virus (16). Several mouse strains were previously claimed to differ in resistance to SeV infection (9, 11, 13–15, 23). Subsequent studies based on those phenotypic differences generated an array of hypotheses about associated characteristics (sex, coat color) or underlying gene polymorphisms (interferon, H-2 haplotype, mucociliary transport, or Sas-1 loci), none of which were finally confirmed (9, 11, 13–15). To evaluate the possibility that using other strains could offer new avenues to identify the genes involved, we compared SeV infections among six strains that were deliberately chosen because they originated from different lineages, as deduced from known genealogical and phylogenetic data (6).
As respiratory pattern/function examination in the mouse by plethysmography yields results far more sensitive and informative than those other techniques can afford when between-strains comparison is the aim (17, 19, 41), we analyzed SeV resistance/susceptibility phenotypes during the first week after infection by simultaneously implementing double-chamber plethysmography, histological evaluation, and lung virus yield assaying techniques. Overall, this approach revealed a pattern of continuous variation, with resistant, intermediate, and susceptible strains, which would be suggestive of a complex trait controlled by several genes.
MATERIALS AND METHODS
Twenty mice from each of six different inbred strains (BALB/cBy, DBA/2, 129Sv, SJL, C3H/HeN, and C57BL/6J) were inoculated with SeV. Respiratory pattern/function values were measured daily in 10 mice from each strain, starting 1 day before inoculation (bi) and going up to 7 days postinoculation (pi). At selected time intervals, a subset of each strain-specific cohort was euthanized for histological examination and lung viral titer quantitation. Precise assignment of mice subsets to the different data collection procedures is presented in Table 1.
Mice, virus, and inoculation.
The experiments were conducted with specific pathogen-free 14-wk-old female mice obtained from Charles River Laboratories. The mice were made familiar with the experimental environment by placing them into the plethysmograph for 15 min a day, starting 3 days bi. Housing, inoculation, data collection, and euthanasia procedures complied with the National Institutes of Health guidelines, and the experimental protocol was approved by the Bioethics Committee of the University of Liège. SeV/52 strain (American Type Culture Collection no. VR-105) was first propagated in 11-day-old chick embryos for 48 h. The harvested allantoic fluid titrated 1/1024 hemagglutination by hemadsorption of guinea pig erythrocytes. It was then diluted to 10−5 in MEM, aliquoted, and stored at −80°C to serve as inoculum. Randomly selected aliquots yielded highly reproducible titers on LLC-MK2 cells, amounting to ∼2 × 104 PFU/ml. The inoculation procedure consisted of slowly instilling 50 μl of the viral suspension (i.e., ∼103 PFU) into the nostrils of the anesthetized mouse maintained in a vertical position (35 mg/kg pentobarbital sodium ip).
Examination of respiratory function and virus yields.
Respiratory pattern/function (RPF) values (Table 2) were measured in 10 mice per strain with the two-chambered, whole body plethysmograph devised by Buxco (model no. PLY-3351), using the practical procedures, quality controls, raw data processing, and respiratory flow curve analysis that were validated before in the laboratory (19). At selected time intervals (5, 6, and 7 days pi), some mice (5, 5, and 10, respectively) were overdosed with pentobarbital sodium and were exsanguinated by cutting the renal artery. The right lung was weighed, homogenized in 10% PBS, and clarified (3,000 g for 10 min), and the supernatant was used for virus titration by plaque assays on LLC-MK2 cells. Once quasiconfluent monolayers grown in 24-well plaques were obtained, the wells were filled with 200 μl of successive tenfold dilutions of the lung homogenates. After the viral suspensions were left to adsorb for 1 h at 37°C, the wells were washed twice with PBS and the cell monolayers were covered with 1 ml of 0.6% agarose (wt/vol) in MEM to which BSA (0.1% wt/vol) and trypsin (10 μg/ml) had been incorporated. After incubation at 37°C for 7 days, the agar overlay was removed, and the remaining cells were stained with crystal violet. The plaques were counted, and the viral titer per unit of right lung weight was calculated. Distribution of SeV antigens was sought by fluorescence microscopy after dewaxed sections were sequentially incubated with anti-SeV antiserum and FITC-conjugated anti-rabbit IgG antibody (Molecular Probes).
Assessment of lung lesion severity.
At the same time, the left lung was inflated with 4% formalin under a pressure of ∼3 kPa for 24 h and then processed along with pieces of nasal turbinates/trachea as for routine histopathological examination. The lung slides were examined independently by two individuals who were never aware of the strain considered. Lesions were graded from “N” (normal) to “+++”. Scattered single inflammatory cells in the alveolar spaces were given a value of +; dispersed clusters of inflammatory cells were given a value of ++; and dispersed confluent areas of inflammatory cells and various amounts of cell debris were given a value of +++. Interstitium cell infiltration was scored + if only small foci of mononuclear cell infiltration were observed, a score of ++ signified that those foci were multifocal and moderate in extension, and finally a score of +++ was assigned when the interstitium was infiltrated by mononuclear cells in a diffuse manner. Lymphoid accumulations were scored + if they were focal and caused no thickening of the submucosa, ++ if they caused mild thickening, and +++ if they were extensive, complete, and resulted in substantial thickening. A score of + for airway epithelial changes indicated hyperplasia without degenerative lesions; a score of ++ signified areas of moderate deciliation and mild exudates; and finally a score of +++ indicated large areas of degeneration alternating with necrosis and severe airway exudates.
Within each strain, consecutive means of RPF values were analyzed by performing the general linear model procedure. Comparisons with day 1 bi were then made by using Dunnett's post hoc test. Effect of strain on log10 means of lung virus titers was analyzed by one-way analysis of variance, and between-strain pairwise comparisons were then made by using Bonferroni's post hoc test. Comparisons that yielded P values of <0.05 were considered statistically significant. Finally, titer-against-day linear regressions were made within each strain and whenever significant yielded a slope value representing the strain-specific lung viral clearance. All statistical analyses were performed with Minitab Release 14 (Minitab statistical software).
All strain- and day-specific values (means ± SD) are presented either in Tables 3–7,Figs. 1–4, or as data supplements (Tables S1-S9; supplemental data for this article may be found at http://ajplung.physiology.org/cgi/content/full/00240.2005/DC1). A synopsis of major strain-specific responses to SeV is also presented in Table 8. None of the 120 inoculated mice died spontaneously during the experiment. As far as the body weight (BW) was concerned, three trends were distinguishable (Fig. 1). The first (129Sv and DBA/2) was a gradual decrease in BW from day 2 (DBA/2) or day 4 (129Sv), which peaked on day 7 with a weight loss of 20%. The second (C57BL/6) was a gradual decrease in BW from day 4, which also peaked on day 7, but with a loss of only 15%. The third (C3H/HEN, SJL, and BALB/c) was stable BW with a slight decrease of ∼5% at day 7.
Regarding the respiratory pattern, each strain reacted in a specific manner to infection (Tables 4, 5, and S6). In 129Sv, the minute volume (MV) remained stable throughout the experiment, but breathing was faster [respiratory rate (RR) > 150% RR bi] and shallower [tidal volume (TV) < 70% TV bi] on and after day 3 pi. In DBA/2, MV evolution was bell shaped: increase from day 4 to day 6 (+30%) and return to normal on day 7. The MV increase was achieved through faster (RR > 175% RR bi) and shallower (TV < 75% TV bi) breathing, and the decrease through lowering of RR without restoration of the initial TV. In C57BL/6, changes also displayed a bell-shaped evolution, with MV remaining at 130% from day 4 to day 6, then decreasing to 115% on day 7, also through faster and shallower breathing. In C3H/HeN, MV was elevated on and after day 4 (∼135% MV bi) and remained at that level until the end of the experiment (day 7). In a manner specific to that strain, hyperpnea was achieved by increasing RR (> 130% RR bi), but with no TV decrease. In SJL, the MV was also increased on and after day 4, but less so (∼110% MV bi), and this level was maintained until day 7. This increase was also achieved through faster and shallower breathing. Finally, in BALB/c, the MV level remained stable throughout the experiment, via a stable RR/TV combination.
When the compartmentation of the duration of each breathing cycle into its component inspiratory time (TI) and expiratory time (TE) was examined [via duty cycle (%Ti), Table S3], significant differences became apparent between strains: decrease of %Ti in 129Sv (on and after day 4), DBA/2 (on day 7), and C57BL/6 (on and after day 2) against stability in C3H/HeN, SJL, and BALB/c.
Focusing on expiration, a qualitative examination of the shape of the breathing waves suggested considerable differences between strains at the time of the final phase of expiration (Fig. 2). For this reason, in addition to the traditional expiration parameters [TE and peak expiratory flow (PEF), Tables S2 and S5], it was necessary to be able to study the evolution of the initial (TE30%, Table S7) and terminal (TE5%, Table S8) phases of expiratory emptying as well as their respective weight in the determination of TE (TE30%/TE, Table 6). With the exception of BALB/c, in which all expiratory parameters remained stable throughout the experiment, TE and TE5% evolved in a parallel and bell-shaped manner (first shortening, then lengthening). More specifically, in 129Sv and DBA/2, TE decreased sharply from day 3 to day 6, then lengthened on day 7, without, however, recovering its initial value. TE5% evolved in the same manner but recovered its initial value in DBA/2 (∼98% of TE5% bi), whereas this was greatly exceeded in 129Sv (∼146% of TE5% bi). In C57BL/6, C3H/HEN, and SJL, a shortening of TE and TE5% was observed and maintained up to the end of the experiment (day 7). Conversely, no parallelism was detected between the evolution of TE30%, which decreased gradually, and the evolution of TE or TE5%. The evolution of the disease, therefore, included recompartmentation of expiration, variable according to the strain, between its initial and terminal components. The TE30%/TE parameter provides a good illustration, with total stability throughout the experiment in C3H/HeN, SJL, and BALB/c, a decrease of ∼15% in C57BL/6 and DBA/2 on day 7, and a decrease of 20% (day 6) and then 30% (day 7) in 129Sv. In short, the times measured on the expiratory wave (TE, TE30%, and TE5%) made it possible to distinguish objectively five “expiratory” entities: 129Sv (shortening of TE5%, then lengthening beyond initial duration with major decrease in TE30%/TE), DBA/2 (shortening then restoration of initial TE5%, with decrease in TE30%/TE), C57BL/6 (shortening maintained but decrease in TE30%/TE on day 7), C3H/HeN and SJL (shortening maintained and TE30%/TE not affected), and BALB/c (expiratory parameters not affected). The PEF did not enable differences to be detected between strains.
Regarding inspiration [TI and peak inspiratory flow (PIF), Tables S1 and S4], four trends appeared. In 129Sv, DBA/2, and C57BL/6, the TI decreased regularly from day 3 (<80% TI bi) to day 7 (<60% TI bi), with a parallel increase in the PIF. It should be noted that TI was already decreasing on day 1 and day 2 in C57BL/6, which was not the case for the other strains. In C3H/HEN and SJL, the TI decreased from day 3 (∼90% of TI bi) to day 4 and remained stable thereafter (∼77% of TI bi); however, they exhibited different PIFs, with an increase until day 4 and maintenance thereafter for C3H/HeN, and continued stability in SJL. Finally, BALB/c distinguished itself by the fact that TI and PIF remained stable.
The changes in %Ti referred to above were therefore due to lesser shortening or to a lengthening of TE in 129Sv and DBA/2, whereas they were due to greater shortening of TI in C57BL/6. In C3H/HeN and SJL, TI and TE decreased in the same proportions, which left %TI unchanged. Clearly, C57BL/6 distinguished itself from the other strains by its propensity to speed up inspiration far more throughout the pathological process, despite an MV level, RR/TV combination, and expiratory characteristics that were never extreme.
As far as respiratory mechanics were concerned, and in particular specific airway resistance (Sraw), three categories became apparent (Table 7). In 129Sv, DBA/2, and SJL, Sraw increased on and after day 3 and then stabilized at ∼150% of its initial value. In C57BL/6 and C3H/HeN, the evolution was the same, but Sraw stabilized at ∼125% of its initial value. Finally, in BALB/c, there was no detectable change in the Sraw. It should be noted that no parallels were identified between the evolution of enhanced pause (Penh) and that of Sraw and that, in the case of C3H/HeN, for instance, experimental infection was accompanied by a decrease in Penh, which is somewhat surprising for a parameter that is supposed to represent resistance.
Nasal turbinates and tracheae displayed very subtle changes, consisting of degeneration of individual epithelial cells and infiltration of some neutrophils in the lamina propria. Three histological profiles were clearly identified on the basis of quantitative and qualitative criteria (Table 2 and Fig. 3). There was little difference between the first histological profile (BALB/cBy) and the normal morphology of murine lungs. The airways almost never contained exudate (60% grade N), and their epithelium appeared generally intact, with the exception of a few isolated losses of ciliation and a few areas of epithelial hyperplasia, slight but definite (30% grade +). The lamina propria contained a few noncoalescent foci of infiltration by mononucleated cells (60% grade + and 20% grade ++). All alveolar spaces were empty, with the exception of a few macrophages, the density of which was comparable to that observable in a healthy lung. In the interstitium, cell density appeared slightly elevated, multifocally. The histopathological diagnosis most compatible with these observations was slight mononuclear broncho-bronchiolitis.
The second histological profile was shared by strains SJL, C3H/HEN, C57BL/6, and DBA/2. The lumen of the airways occasionally (C57BL/6 and C3H/HEN, both with 40% of the lungs with a score at least grade ++) or systematically (SJL and DBA/2, both with more than 80% of the lungs scoring at least grade ++) contained an exudate identical to that found in the alveolar spaces. This was a mixture of cell debris, epithelial cells, pleiomorphic lymphoid cells, neutrophils, and morphologically altered macrophages (cytoplasmic vacuolation, pycnosis, caryorhexis). The epithelial lining always exhibited large areas of deciliation alternating with degeneration, necrosis, desquamation, and marked hyperplasia with uneven thicknesses and cell arrangements. The lamina propria was infiltrated by numerous monocytoid or lymphoid cells with round nuclei, which generated so-called “peribronchial/olar and perivascular cuffing” images, in SJL. Fifty percent were grade ++. In C3H/HEN and DBA/2, at least 70% of the lungs were scored equal or above grade ++, and in C57BL/6, they were complete and very thick in 80% of the lungs, and the more obvious the larger the airways concerned. Most of the alveolar spaces were optically empty, but often (∼70% of individuals in SJL, C3H/HEN, and C57) or systematically (DBA/2, 100% with a grade ≥++), groups of alveoli were observed that contained the exudate described above. In general, these alveolar clusters were adjacent to an airway and, in C57BL/6 and DBA/2, tended to merge, giving the impression of conquering the more peripheral spaces gradually. Overall, the number of alveoli concerned varied considerably between strains, from a few dozen in SJL and C3H/HEN to more than one-third of the lung section area in DBA/2. The lung interstitium, when examined away from the areas of alveolitis, was, with the exception of SJL, multifocally (C3H/HEN and C57BL/6) or diffusely (DBA/2 40% grade ++ and 20% grade +++) infiltrated by cells with round nuclei of the monocytoid or lymphoid type. The density of the alveolar macrophages was normal, except in DBA/2, in which it was elevated. The histopathological diagnoses most compatible with these observations were severe necrotizing and purulent broncho-bronchiolitis and multifocal alveolitis.
The third histological profile concerned strain 129Sv. Until day 7, the airways never contained exudate. The changes to the epithelium and lamina propria contrasted radically with the previous profiles. There was little necrosis or desquamation of the epithelium (30% grade ++), but diffuse and severe hyperplasia (70% grade +). The peribronchial/olar and perivascular cuffing was thicker, 90% was scored at least ++, and of those, 70% were of grade +++. The specificity of the histological profile of 129Sv mainly related to the distal areas. The alveolar spaces never contained the exudate described above, and therefore no neutrophils, but contained an extremely high number of macrophages, of which only a small proportion exhibited regressive alterations (cytoplasmic vacuolation, pycnosis, caryorhexis). Certain areas were noteworthy for the presence of groups of alveoli completely filled with closely packed macrophages. Moreover, the interstitium was severely infiltrated by cells with round nuclei in a diffuse manner. The histopathological diagnoses most compatible with these observations were severe bronchial/olar epithelial hyperplasia and diffuse monocytoid interstitial pneumonia.
The titers (PFU/g of lung) obtained on days 5, 6, and 7 are presented in Fig. 4. The 129Sv strain combined the highest titer on day 5 and the slowest day 5 > day 7 clearance (with C57BL/6). Conversely, BALB/c combined the lowest day 5 titer and the fastest day 5 > day 7 clearance (with SJL). Between these extremes, the lungs of C3H/HEN and SJL generated a higher day 5 titer than BALB/c but, like the latter, exhibited fast clearance. In C57BL/6, the day 5 titer was comparable to that of SJL and C3H/HeN, but clearance was very slow, comparable to that of 129Sv. Finally, in DBA/2, the day 5 titer was high and the clearance better than in 129Sv and C57BL/6. Topological distribution of viral antigens within the lungs also generated very different results according to the strain (Fig. 3). The 129Sv strain was distinguished by the presence of virus-positive bronchial/olar epithelial cells, type-1 and -2 pneumocytes, and macrophages homogeneously distributed throughout the lungs. In C57BL/6 and DBA/2, the virus-positive cells occupied an accumulated area of ∼20% and ∼30%, respectively, of the sample section area, the distribution of which roughly overlapped with the areas of alveolitis identified histologically. In C3H/HeN, SJL, and BALB/c, only a few virus-positive bronchial epithelial cells were identified, and nothing in BALB/c on day 7.
One hundred twenty mice from six different inbred strains were nasally inoculated with a standardized suspension containing SeV. This inoculation caused a viral respiratory illness, as demonstrated by the changes in RPF values, respiratory histology, and the pulmonary viral titers recorded 5–7 days thereafter. Significant differences appeared between strains, to the extent that a wide phenotypic spectrum was established, for clinical (asymptomatic vs. symptomatic) and tissue responses (necrosis vs. hyperplasia), cell infiltration (neutrophilic vs. macrophagic), viral peak titer (high, medium, low), clearance (fast, slow, or absent), and viral tropism.
The priority was to achieve a level of standardization that would in fine enable the differences in the severity of the disease induced to be explained in terms of genetic background, thus making it possible, at least in theory, to designate a “resistant” and a “susceptible” strain at the end of the experiment. Regarding the pathogen, all the mice were infected with an identical number of PFUs from the same stock of a well-characterized viral strain, diluted in an identical volume of MEM. Preliminary experiments had duly established that the volume instilled in the nostrils using the selected procedure systematically caused diffuse deposition in the lungs while avoiding sneezing at the time of inoculation (which would have reduced the effective viral load) and the presence of changes in the breathing pattern during the following hours. As far as the hosts were concerned, all the mice were female, aged 14 wk, and sero-negative for SeV and the most common murine pathogens. Thus the functional parameters reported were unaffected by changes related to gender and somatic growth (19). The method chosen (double-chamber plethysmography) has been revealed to be a trustful and sensitive technique to evaluate the severity of respiratory diseases. Before inoculation, it generated reproducible baseline physiological measurements on the same animal subjects. After inoculation, it generated reliable and repeated ventilatory and mechanics changes that were well correlated with histological and virological data. The experimental protocol as a whole generated an infectious model that was neither too benign (which would have reduced experiment sensitivity) nor too severe (which would have gone against the experiment's specificity and hindered the statistical analysis of the results). The initial decision to restrict virological examination to days 5 to 7 pi was based on available data as well as on our own preliminary studies (data not shown) that had established that peak viral titers were never reached before the fifth day pi (5, 13, 14), whatever the strain. Here, the peak titers were recorded on day 5 in all strains and declined thereafter, thus suggesting that shifts in the time course of viral replication among strains could indeed be interpreted in terms of differences in innate (peak) or adaptive (clearance) immune responses. Each of the methods used to monitor the induced disease generated objective categories, ranging from “favorable” to “unfavorable.” The combination of the functional, histological, and virological categories (Table 8) led to a classification of the strains in terms of resistance to the disease induced by SeV: BALB/c > SJL > C3H/HeN > C57BL/6 > DBA/2 > 129Sv.
A specific access to the respiratory system was given by the monitoring of plethysmographic values, especially when considering the descriptive parameters of global breathing and expiration. In general, inoculation of SeV resulted in a combination of tachypnea, hyperpnea, and alteration of expiratory emptying. As inspiratory duration, peak flow, and flow trace were never lengthened, blunted, or reshaped, it could be deduced that the induced pathological process did not involve increased resistance to airflow in the upper airways (18). On the favorable side, the breathing pattern and mechanics (Sraw) of BALB/c remained unchanged throughout the experiment, which shows that the respiratory disease caused by SeV remained totally asymptomatic. SJL and C3H/HeN differed from BALB/c in that they developed compensatory hyperpnea via faster breathing, the level of which was maintained up to the end of the experiment. Sraw also increased, which points to a diminished cross-section area of the airways (19, 31). To the best of our knowledge, the partitioning of Sraw between the different segments of the mouse respiratory tree is not available yet, which renders reliable interpretation in terms of location of the source of the problem very speculative. However, it was recently shown that nasal obstruction alone can substantially increase Sraw in mice (27, 28). Because turbinate histology did not differ between BALB/c, SJL, and C3H/HeN and according to the fact that the mouse breathing pattern response typical of nasal irritation (slowing down of the beginning of expiration) or narrowing was never observed (3, 8, 18, 43), it is tempting to conclude that the increased Sraw in the latter two results from tracheobronchial alterations rather than from superior nasal congestion and/or thickening of the mucosa.
In C57BL/6, all functional parameters evolved in a manner similar to that observed in SJL and C3H/HeN until day 6 when two additional events occurred: 1) the terminal phase of expiratory emptying lengthened (TE5%) and 2) the level of hyperpnea (MV) was no longer maintained because of the increased time spent to exhale. In this connection, the expiratory flow profiles recorded from C57BL/6 (Fig. 2) strikingly resemble those known as being typical of deep pulmonary irritation/inflammation (3, 8, 27–29, 36, 43). Together with the fact that, in this study, only those strains with lengthened TE5% (C57BL/6, DBA/2, and 129/Sv) were subsequently shown to develop severe alveolar alterations, it is suggested that SeV infection resulted in deeper and/or more extensive lesions in those strains than in SJL and C3H/HeN. It therefore appears obvious that the nonmaintenance of MV does not mean improvement of the respiratory function, but rather the inability to maintain the level of polypnea in the face of the braking of lung emptying.
In DBA/2 and 129Sv, the evolution of the functional parameters is qualitatively comparable to that for C57BL/6, which therefore generates a comparable interpretation. However, quantitatively speaking, RPF value changes were far more severe. The intensity of respiratory dysfunctions culminated in 129Sv, with a collapse in TE30%/TE on day 6 and a lengthening of the TE5% such (146%) that it greatly exceeded its initial value before infection, despite the concomitant tachypnea. Regarding the breathing pattern as a whole, DBA/2 was similar to C57BL/6, but with more marked decompensation, as the MV returned to its preinfection level on day 7, which had never been the case in C57BL/6. In 129Sv, another specific characteristic should be noted in this respect, as this is the only strain that was unable to increase its MV, even at the beginning of the infectious process when the impediment of breathing did not yet appear to have occurred. Indeed, the frequency response was less marked in 129Sv than in DBA/2 (RR, ∼150 vs. ∼175%) whereas, conversely, DBA/2 maintained a better TV (∼75 vs. ∼70%) despite this higher RR. This specificity appears difficult to interpret, as it may be caused by differences in central breathing control, in the afferent perception of imbalances (chemosensors, stretch receptors), or in the efferent neuromechanical transduction.
On the functional level also, two breathing strategies should be noted that are probably unrelated to any degree of resistance/susceptibility to SeV infection. In C57BL/6, the reaction to inoculation includes a different management of inspiration/expiration compartmentation, with a decrease in %TI throughout the experiment. This decrease does also become apparent later in DBA/2 and 129Sv but is due to a lengthening of the TE, whereas in C57BL/6, this decrease is always due to a shortening of the TI. This observation suggests that C57BL/6 needs to use a different strategy to increase its MV, possibly because its thoracoabdominal mechanics are different, apparently less compliant (22, 34, 37, 41). Moreover, C3H/HeN is the only strain that increases its MV by relying exclusively on frequency, i.e., without decreasing the TV. To our knowledge, this strategy is quite specific, as the general rule observed to date on all animals is that TV decreased when RR increased. It may be supposed that this strategy may also be linked to the thoracopulmonary structure specific to this strain, which has larger pulmonary volumes and higher pulmonary compliance (22, 34, 37, 41).
Although nonspecific, the monitoring of live weight enabled two (at least) main categories to be determined: favorable for BALB/c, SJL, and C3H/HeN, which did not lose weight, and fairly unfavorable for C57BL/6 and especially for DBA/2 and 129Sv, the BW loss of which exceeded 20% in 7 days.
In general, SeV infection caused epithelial progressive (hyperplasia) and regressive (deciliation, degeneration, necrosis, and exfoliation) lesions and inflammatory infiltrations. Although no obvious between-strain difference was recorded in turbinates and tracheae, lung lesions were clearly strain specific, suggesting that the differences recorded in the RPF values are not attributable to the upper airways. In BALB/c, hyperplasia was discreet, regressive lesions were extremely rare, and exudates and inflammatory infiltrations were absent, which suggests the presence of viral replication being repressed by powerful resistance mechanisms. In 129Sv, epithelial hyperplasia and diffuse interstitial lymphoid and monocytoid infiltration predominated, whereas exudates, granulocytes, and regressive lesions were absent. These observations suggest intense epithelial viral replication specifically recruiting mononucleated cells. In SJL, C3H/HeN, C57BL/6, and DBA/2, all aspects of infection were present, with an earlier apparition of necrosis and exudates in SJL and C3H/HeN than in C57BL/6 and DBA/2. The systematically concomitant presence of neutrophils and exudates in the bronchiolar/alveolar lumina, and of necrosis and exfoliation lesions, suggests that there may a be cause to effect a link between both phenomena. In this case, it may be that in 129Sv, granulocyte infiltration occurs after day 7 and that the lesions would have been qualitatively similar to those observed to various degrees in SJL, C3H/HeN, C57BL/6, and DBA/2. Moreover, the intensity of lymphoid recruitment in the periphery of the airways varied a great deal from one strain to another. In this area, it has been said that greater local recruitment permits more effective viral eradication (7, 13, 35). However, in this experiment, it is clear that independently of the general histological profile, C57BL/6 and 129Sv exhibited higher peribronchial lymphoid recruitment, although they displayed the slowest viral clearance. Conversely, SJL, which is one of the two most resistant strains, was characterized by thin, incomplete cuffing and therefore by lower lymphoid recruitment. These observations invalidate the hypothesis according to which “thick peribronchial cuffing protects against the virus.”
The results show that 1) the strain affects the maximum viral titer, 2) the strain affects the decrease in viral titer from day 5 to day 7, 3) there is no absolute correlation between the maximum titer and the decrease in viral titers, and 4) SeV pulmonary pantropism correlates with highest severity. In general, with two exceptions (14, 15), the data and interpretations available in the literature led most authors to conclude that the maximum titer was not affected by the strain, whereas mortality and lesion and symptom severity were (9, 13, 23). Conversely, our results suggest that there is indeed a connection between the pulmonary titers and the morphological and clinical characteristics of the infection. To explain this apparent contradiction, it should be noted here that the adjunction of trypsin plays a notorious part in the success of viral replication in cell cultures at the time of titration. Incorporated trypsin cleaves the precursor of SeV glycoprotein F, which activates viral infectivity in a way comparable to that which tryptase secreted from Clara cells assumes in vivo (40, 42, 44). Most of the available studies systematically neglect this aspect, which make comparisons between reported pulmonary viral titers highly questionable (9, 13). On the basis of this methodological criterion, only the titers reported for BALB/c and DBA/2 by Itoh et al. (23) can be compared with those generated here. However, the interpretation of the results of this study remains difficult, on the authors' own admission, as the titrations (similar titers) and immunofluorescence (“there are more SeV antigens in the bronchial epithelium and far more in the alveoli in DBA/2 compared with BALB”) generate contradictory results. The present study thus suggests that resistance is associated to the host capacity to repress the initial multiplication of the virus, to its capacity to eliminate the virus, and to its capacity to restrict the spatial extension of viral multiplication to the epithelium of the airways.
Resistance/susceptibility patterns to SeV in the mouse.
This study has made it possible to characterize several phenotypes relating to the reaction of murine hosts to SeV. BALB/c exhibits a benign and asymptomatic affection of the epithelium of the airways, with no functional impact, generating slight mononucleated cell infiltration, in which viral replication is repressed and the virus swiftly eliminated. SJL, C3H/HeN, C57BL/6, and DBA/2 exhibit moderate (SJL) to severe (DBA/2) lesions of the epithelium of the airways, symptomatic, with resistive-type functional impact, leading to simultaneous granulocyte and mononucleated cell multifocal infiltrations, with regressive lesions, in which viral replication is more (SJL, C3H/HeN, and C57BL/6) or less (DBA/2) repressed and in which the virus is more (SJL and C3H/HeN) or less (C57BL/6 and DBA/2) swiftly eliminated. Finally, 129Sv exhibits a long-term proliferative attack of the epithelium of the airways and alveoli, symptomatic, with resistive and restrictive functional impact, leading to diffuse lymphoid/monocytoid (never granulocytic, at least before day 7) infiltration, in which viral replication is not much repressed and is only slowly eliminated. These observations, made on strains that have been inbred for a long time, suggest plainly that genetic background strongly influences the apparent virulence of SeV and, by extension, of Paramyxoviridae infection.
Candidate resistance genes.
Interestingly, available C3H strains are derived from a mating between DBA and progenitors of BALB/c mice (6). As resistance to SeV appears to depend on genetic background, it is thus expected that C3H strains share one or some alleles with BALB/c that render them more resistant to the virus than DBA. A series of genetic loci have already been identified of which allelic variation underlie susceptibility/resistance patterns to viral infections in mice, i.e., Mx1 for influenza viruses (39), Mx2 for vesicular stomatitis virus and hantaviruses (24, 25), Rmp for mousepox virus (12), Ly49h for cytomegalovirus (10), Bgp1 for mouse hepatitis virus (30), and Flv for flaviviruses (21). These loci are unlikely to play a role here. Indeed, the six strains used had been proven to carry the susceptible alleles at Mx1 and Mx2 loci; BALB/c and DBA/2 share a disease-susceptible phenotype against mouse poxvirus, whereas they significantly differ in resistance here; BALB/c carries the susceptible allele at the Ly49h locus and C57BL/6 the resistant allele, whereas BALB/c is far more resistant to SeV; and none of the strains used carries the resistant allele at the Flv locus. Therefore, only the Bgp1 locus could, at least theoretically, play a role in explaining the results gathered here since the strain SJL combines a resistant phenotype against SeV and mouse hepatitis virus, the latter being associated with a specific allele at this locus. However, the fact that BALB/c displays the most resistant phenotype against SeV while carrying the susceptible allele at the Bgp1 locus along with the contrast between the respiratory specificity of SeV on the one hand and the biliary tract-specific expression of Bgp1 on the other strongly pleads against the hypothesis that Bgp1 could be involved. Thus it is suggested that resistance to SeV is not underlined by the viral resistance loci identified so far. The present study suggests that BALB/c and 129Sv strains should be used in crossing experiments aimed at identifying the genes involved in resistance to SeV by the positional cloning approach.
We are grateful for the scientific expertise provided by Dominique Cassart, Etienne Baise, Michaël Leroy, and Thierry Flandre. We also thank Michaël Sarlet and Grégory Pire for excellent technical skills and enthusiasm.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2005 the American Physiological Society