Limits...
Specific mutations in H5N1 mainly impact the magnitude and velocity of the host response in mice.

Tchitchek N, Eisfeld AJ, Tisoncik-Go J, Josset L, Gralinski LE, Bécavin C, Tilton SC, Webb-Robertson BJ, Ferris MT, Totura AL, Li C, Neumann G, Metz TO, Smith RD, Waters KM, Baric R, Kawaoka Y, Katze MG - BMC Syst Biol (2013)

Bottom Line: Using a new geometrical representation method and two criteria, we show that inoculation concentrations and four specific mutations in VN1203 mainly impact the magnitude and velocity of the host response kinetics, rather than specific sets of up- and down- regulated genes.These kinetic properties imply that time-matched comparisons of 'omics profiles to viral infections give limited views to differentiate host-responses.Moreover, these results demonstrate that a fast activation of the host-response at the earliest time points post-infection is critical for protective mechanisms against fast replicating viruses.

View Article: PubMed Central - HTML - PubMed

Affiliation: Department of Microbiology, University of Washington, Seattle, WA 98195 USA.

ABSTRACT

Background: Influenza infection causes respiratory disease that can lead to death. The complex interplay between virus-encoded and host-specific pathogenicity regulators - and the relative contributions of each toward viral pathogenicity - is not well-understood.

Results: By analyzing a collection of lung samples from mice infected by A/Vietnam/1203/2004 (H5N1; VN1203), we characterized a signature of transcripts and proteins associated with the kinetics of the host response. Using a new geometrical representation method and two criteria, we show that inoculation concentrations and four specific mutations in VN1203 mainly impact the magnitude and velocity of the host response kinetics, rather than specific sets of up- and down- regulated genes. We observed analogous kinetic effects using lung samples from mice infected with A/California/04/2009 (H1N1), and we show that these effects correlate with morbidity and viral titer.

Conclusions: We have demonstrated the importance of the kinetics of the host response to H5N1 pathogenesis and its relationship with clinical disease severity and virus replication. These kinetic properties imply that time-matched comparisons of 'omics profiles to viral infections give limited views to differentiate host-responses. Moreover, these results demonstrate that a fast activation of the host-response at the earliest time points post-infection is critical for protective mechanisms against fast replicating viruses.

Show MeSH

Related in: MedlinePlus

Intersections between the lists of DET and DEP for the strain conditions. (A) and (B) Proportional Euler-diagrams showing the intersections between the lists of differentially expressed transcripts and proteins identified for the strain conditions. Proportional Euler-diagrams visually represent the cardinalities of sets and intersection sets of differentially expressed transcripts or proteins by area-proportional circle graphics. Each list of DE transcripts or proteins is then represented by a circle with a diameter proportional to the cardinality of the list and the overlaps between the circles are proportional to the cardinality of the intersections between the lists. For each strain condition the number of transcripts or proteins found as DE in the host response is indicated as well as the degree of overlap — quantified as the percentage of transcripts or proteins also found as differentially expressed in another condition. (C) Heatmap of the transcriptomic expression values, ratioed to mock-infected samples, for each infected sample. The heatmaps have been restricted to the lists of transcripts found as specific for each viral strain and the different subsets of transcripts specific are indicated. For each set of specific transcripts, hierarchical clustering have been performed and represented using dendrograms. Biological samples have been ordered by strain conditions, sorted by inoculation concentrations and then by increasing days post-infection. (D) Heatmaps showing the statistical over-representation of the canonical pathways based on the lists of transcripts found as differentially expressed (compared to the mock-infected conditions) for each of the 51 biological conditions. This heatmap has been restricted to only display the top canonical pathways over-represented across all the dataset.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3750405&req=5

Figure 2: Intersections between the lists of DET and DEP for the strain conditions. (A) and (B) Proportional Euler-diagrams showing the intersections between the lists of differentially expressed transcripts and proteins identified for the strain conditions. Proportional Euler-diagrams visually represent the cardinalities of sets and intersection sets of differentially expressed transcripts or proteins by area-proportional circle graphics. Each list of DE transcripts or proteins is then represented by a circle with a diameter proportional to the cardinality of the list and the overlaps between the circles are proportional to the cardinality of the intersections between the lists. For each strain condition the number of transcripts or proteins found as DE in the host response is indicated as well as the degree of overlap — quantified as the percentage of transcripts or proteins also found as differentially expressed in another condition. (C) Heatmap of the transcriptomic expression values, ratioed to mock-infected samples, for each infected sample. The heatmaps have been restricted to the lists of transcripts found as specific for each viral strain and the different subsets of transcripts specific are indicated. For each set of specific transcripts, hierarchical clustering have been performed and represented using dendrograms. Biological samples have been ordered by strain conditions, sorted by inoculation concentrations and then by increasing days post-infection. (D) Heatmaps showing the statistical over-representation of the canonical pathways based on the lists of transcripts found as differentially expressed (compared to the mock-infected conditions) for each of the 51 biological conditions. This heatmap has been restricted to only display the top canonical pathways over-represented across all the dataset.

Mentions: For each biological condition, differentially expressed transcripts (DET) and proteins (DEP) compared to the mock-infected samples were identified and the overlap between the lists of DET or DEP of each virus were compared. Figure 2A and B provide visual representations of the intersections between the lists of DET and DEP for the strain conditions using proportional Euler-diagrams [23]. There were large degrees of overlap (defined here as the percentage of DET or DEP also found in another condition) between the strain conditions, between 69.41% - 99.62% and 85.94% - 96.98%, DET and DEP, respectively. The VN1203-PB1F2del strain condition had the smallest degree of overlap at the transcriptomic (69.41%) level, and a relatively small degree at the proteomic (87.95%) levels. Notably, at 7 dpi, only one transcriptomic profile was available for the VN1203-PB1F2del strain condition inoculated at the 104 PFU concentration. Hence, with only one sample available, the statistical identification of DET lead to the identification of a fraction of false-positive, explaining this lowest degree of overlap observed for this strain condition at the transcriptomic level. The VN1203-WT strain condition also showed a relatively small degree of overlap at the transcriptomic (85.55%) level; and the VN1203-NS1trunc strain condition showed the smallest degree of overlap at the proteomic (85.94%) level compared to the other strain conditions. It should be noted that the strain conditions with the lowest degrees of overlap corresponded to the same strain conditions that triggered the largest amount of DET and DEP. Similar large amounts of overlap have been identified between the dosage conditions, with degrees of overlap ranging from 68% to 98.81% for the transcriptome and from 89.87% to 99.89% for the proteome. Despite differences in clinical manifestation of the disease, these results suggest that the specific mutations in VN1203 examined herein resulted in the induction of similar groups of transcripts and proteins compared to VN1203-WT, implying that magnitude and/or the kinetics of dysregulation of these overlapping genes might differentiate the viruses. Figure 2C is a heatmap of the transcript expression values, ratioed to the mock-infected samples, of each infected sample of our dataset. This heatmap has been restricted to the lists of transcripts specific to each strain condition (i.e. transcripts found as differentially expressed within one viral strain but not in the others). We cannot distinguish any particular patterns or sets of transcript specific to any viral strain based on these transcript expression values. The transcript or protein subsets specific to each viral strain are the consequences of small variations in the host response, detected by the statistical procedures, that have no specific biological meaning.


Specific mutations in H5N1 mainly impact the magnitude and velocity of the host response in mice.

Tchitchek N, Eisfeld AJ, Tisoncik-Go J, Josset L, Gralinski LE, Bécavin C, Tilton SC, Webb-Robertson BJ, Ferris MT, Totura AL, Li C, Neumann G, Metz TO, Smith RD, Waters KM, Baric R, Kawaoka Y, Katze MG - BMC Syst Biol (2013)

Intersections between the lists of DET and DEP for the strain conditions. (A) and (B) Proportional Euler-diagrams showing the intersections between the lists of differentially expressed transcripts and proteins identified for the strain conditions. Proportional Euler-diagrams visually represent the cardinalities of sets and intersection sets of differentially expressed transcripts or proteins by area-proportional circle graphics. Each list of DE transcripts or proteins is then represented by a circle with a diameter proportional to the cardinality of the list and the overlaps between the circles are proportional to the cardinality of the intersections between the lists. For each strain condition the number of transcripts or proteins found as DE in the host response is indicated as well as the degree of overlap — quantified as the percentage of transcripts or proteins also found as differentially expressed in another condition. (C) Heatmap of the transcriptomic expression values, ratioed to mock-infected samples, for each infected sample. The heatmaps have been restricted to the lists of transcripts found as specific for each viral strain and the different subsets of transcripts specific are indicated. For each set of specific transcripts, hierarchical clustering have been performed and represented using dendrograms. Biological samples have been ordered by strain conditions, sorted by inoculation concentrations and then by increasing days post-infection. (D) Heatmaps showing the statistical over-representation of the canonical pathways based on the lists of transcripts found as differentially expressed (compared to the mock-infected conditions) for each of the 51 biological conditions. This heatmap has been restricted to only display the top canonical pathways over-represented across all the dataset.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3750405&req=5

Figure 2: Intersections between the lists of DET and DEP for the strain conditions. (A) and (B) Proportional Euler-diagrams showing the intersections between the lists of differentially expressed transcripts and proteins identified for the strain conditions. Proportional Euler-diagrams visually represent the cardinalities of sets and intersection sets of differentially expressed transcripts or proteins by area-proportional circle graphics. Each list of DE transcripts or proteins is then represented by a circle with a diameter proportional to the cardinality of the list and the overlaps between the circles are proportional to the cardinality of the intersections between the lists. For each strain condition the number of transcripts or proteins found as DE in the host response is indicated as well as the degree of overlap — quantified as the percentage of transcripts or proteins also found as differentially expressed in another condition. (C) Heatmap of the transcriptomic expression values, ratioed to mock-infected samples, for each infected sample. The heatmaps have been restricted to the lists of transcripts found as specific for each viral strain and the different subsets of transcripts specific are indicated. For each set of specific transcripts, hierarchical clustering have been performed and represented using dendrograms. Biological samples have been ordered by strain conditions, sorted by inoculation concentrations and then by increasing days post-infection. (D) Heatmaps showing the statistical over-representation of the canonical pathways based on the lists of transcripts found as differentially expressed (compared to the mock-infected conditions) for each of the 51 biological conditions. This heatmap has been restricted to only display the top canonical pathways over-represented across all the dataset.
Mentions: For each biological condition, differentially expressed transcripts (DET) and proteins (DEP) compared to the mock-infected samples were identified and the overlap between the lists of DET or DEP of each virus were compared. Figure 2A and B provide visual representations of the intersections between the lists of DET and DEP for the strain conditions using proportional Euler-diagrams [23]. There were large degrees of overlap (defined here as the percentage of DET or DEP also found in another condition) between the strain conditions, between 69.41% - 99.62% and 85.94% - 96.98%, DET and DEP, respectively. The VN1203-PB1F2del strain condition had the smallest degree of overlap at the transcriptomic (69.41%) level, and a relatively small degree at the proteomic (87.95%) levels. Notably, at 7 dpi, only one transcriptomic profile was available for the VN1203-PB1F2del strain condition inoculated at the 104 PFU concentration. Hence, with only one sample available, the statistical identification of DET lead to the identification of a fraction of false-positive, explaining this lowest degree of overlap observed for this strain condition at the transcriptomic level. The VN1203-WT strain condition also showed a relatively small degree of overlap at the transcriptomic (85.55%) level; and the VN1203-NS1trunc strain condition showed the smallest degree of overlap at the proteomic (85.94%) level compared to the other strain conditions. It should be noted that the strain conditions with the lowest degrees of overlap corresponded to the same strain conditions that triggered the largest amount of DET and DEP. Similar large amounts of overlap have been identified between the dosage conditions, with degrees of overlap ranging from 68% to 98.81% for the transcriptome and from 89.87% to 99.89% for the proteome. Despite differences in clinical manifestation of the disease, these results suggest that the specific mutations in VN1203 examined herein resulted in the induction of similar groups of transcripts and proteins compared to VN1203-WT, implying that magnitude and/or the kinetics of dysregulation of these overlapping genes might differentiate the viruses. Figure 2C is a heatmap of the transcript expression values, ratioed to the mock-infected samples, of each infected sample of our dataset. This heatmap has been restricted to the lists of transcripts specific to each strain condition (i.e. transcripts found as differentially expressed within one viral strain but not in the others). We cannot distinguish any particular patterns or sets of transcript specific to any viral strain based on these transcript expression values. The transcript or protein subsets specific to each viral strain are the consequences of small variations in the host response, detected by the statistical procedures, that have no specific biological meaning.

Bottom Line: Using a new geometrical representation method and two criteria, we show that inoculation concentrations and four specific mutations in VN1203 mainly impact the magnitude and velocity of the host response kinetics, rather than specific sets of up- and down- regulated genes.These kinetic properties imply that time-matched comparisons of 'omics profiles to viral infections give limited views to differentiate host-responses.Moreover, these results demonstrate that a fast activation of the host-response at the earliest time points post-infection is critical for protective mechanisms against fast replicating viruses.

View Article: PubMed Central - HTML - PubMed

Affiliation: Department of Microbiology, University of Washington, Seattle, WA 98195 USA.

ABSTRACT

Background: Influenza infection causes respiratory disease that can lead to death. The complex interplay between virus-encoded and host-specific pathogenicity regulators - and the relative contributions of each toward viral pathogenicity - is not well-understood.

Results: By analyzing a collection of lung samples from mice infected by A/Vietnam/1203/2004 (H5N1; VN1203), we characterized a signature of transcripts and proteins associated with the kinetics of the host response. Using a new geometrical representation method and two criteria, we show that inoculation concentrations and four specific mutations in VN1203 mainly impact the magnitude and velocity of the host response kinetics, rather than specific sets of up- and down- regulated genes. We observed analogous kinetic effects using lung samples from mice infected with A/California/04/2009 (H1N1), and we show that these effects correlate with morbidity and viral titer.

Conclusions: We have demonstrated the importance of the kinetics of the host response to H5N1 pathogenesis and its relationship with clinical disease severity and virus replication. These kinetic properties imply that time-matched comparisons of 'omics profiles to viral infections give limited views to differentiate host-responses. Moreover, these results demonstrate that a fast activation of the host-response at the earliest time points post-infection is critical for protective mechanisms against fast replicating viruses.

Show MeSH
Related in: MedlinePlus