Background
Huntington's disease (HD) is an autosomal-dominant neurodegenerative disorder which is caused by polyglutamine (polyQ) expansion in the amino-terminus of huntingtin (Htt). Htt is a soluble protein of 3,144 amino acids that has no sequence homology with other proteins. Except for the extreme amino-terminus, with its adjacent polyQ region and proline-rich segments, the entire ~350-kD protein is predicted to be composed of 36 α-helical HEAT repeats. Increasing evidence indicates that Htt functions as a molecular scaffold that is able to organize a variety of signaling complexes 12. Htt is expressed ubiquitously in humans and rodents, with the highest levels found in CNS neurons and the testes 345. Intracellularly, Htt is associated with various organelles, including the nucleus, endoplasmic reticulum (ER) and Golgi complex 678. This widespread subcellular localization does not facilitate the definition of its function. Hdh is evolutionary conserved – a single copy of the Htt gene is expressed in all vertebrates (from fish to humans) 9. The Htt gene is also present in D. melanogaster genome, but absent in the C. elegans and S. cerevisiae genomes 9. All vertebrate isoforms of Htt, but not Drosophila Htt, contain an amino-terminal polyQ region.
Complete knockout of the mouse Htt gene (Hdh) causes embryonic death before day 8.5 (E8.5, before gastrulation and the formation of the nervous system) 101112. After gastrulation, Htt becomes important for neurogenesis – mice carrying a <50% dose of wild-type Htt display profound malformations of the cortex and striatum 13. Another study has shown that greatly reduced Htt levels are insufficient to support normal mouse development 14. In addition to its function in development, Htt may play a role in the regulation of apoptosis, control of BDNF production, vesicular and mitochondrial transport, neuronal gene transcription, and synaptic transmission (reviewed in 9). Despite all of these efforts and results, the exact function of Htt in cells still remains largely unknown.
In addition to answering an academic question concerning the normal function of Htt, knowledge of its function is important for understanding HD pathogenesis and for the treatment of Huntington's disease (HD). Although the HD mutation is considered to be a "gain of function" mutation, it has been suggested that the loss of normal Htt function might also contribute to the pathogenesis of HD 9. Approaches that are based on reducing mutant Htt expression such as RNA interference 15 and the use of intrabodies 1617 are currently considered to be promising strategies for HD treatment. It is likely that these agents will cause inactivation or impair normal function of both mutant and wild type Htt alleles. One can envision a therapy that combines such Htt-inactivating agents with drugs that restore the function of targets and pathways downstream from wild-type Htt. However, because both normal Hdh function is not known and downstream pathways have not been identified, such a combined therapy approach is not feasible at the moment.
To systematically search for Htt's normal function, we used Hdh +/- 12 and Hdh-floxed mice 18 to generate immortalized mouse embryonic fibroblasts (MEF) which contain a single functional copy of Hdh gene (Hdh-HET) or lack Hdh completely (Hdh-KO). We compared inositol 1,4,5-trisphosphate receptor (InsP3R)-mediated Ca2+ signals in these cells. We then performed a genome-wide gene transcription profiling of Hdh-HET and Hdh-KO MEF cells using microarrays to obtain novel, unique, and unbiased information about the normal function of Htt in fibroblasts, which may contribute to our understanding and treatment of HD.
Results
Generation of Hdh-HET and Hdh-KO MEF cell lines
To generate cell lines lacking Htt expression, we employed a conditional mutagenesis strategy based on the in vitro recombination of an Hdh(flox) allele in cultured fibroblasts that are also carrying either a wild-type (+) or null (standard knock-out) Hdh allele. The Hdh(flox/+) and Hdh(flox/-) fibroblasts were obtained from embryos derived from a cross between Hdh +/- and Hdh-floxed/floxed mice (Fig 1). Primary fibroblasts were prepared and plated separately from each embryo as described in Methods. After two days in culture, the primary fibroblasts from all embryos with identical genotype (Hdh floxed/+ or Hdh floxed/-) were pooled together and transfected with a linearized SV40 plasmid. Transfected cells were then cultured for four to six weeks until immortalized Hdh floxed/+ and Hdh floxed/- mouse embryonic fibroblasts (MEFs) were obtained (Fig 1). To recombine the Hdh floxed allele, immortalized Hdh floxed/+ and Hdh floxed/- MEFs were infected with Lenti-NLS-GFP-Cre virus encoding nuclear-targeted GFP-Cre fusion protein 19 (Fig 1). Using the procedure described above, we generated four Hdh-HET (lines 1,2,3,5) and four Hdh-KO (lines 11, 12, 16, 27) MEF cell lines. The expression of Htt in the generated MEF lines was assessed by Western blotting of whole cell lysates using anti-Htt monoclonal antibody. Quantification of Western blotting data verified similar levels of Htt expression in all 4 Hdh-HET lines (data not shown). Consistent with the genotype of the generated cells, we detected a protein of predicted size (~350 kD) in lysates from the Hdh-HET cells, but not in lysates from Hdh-KO lines (Fig 2). The same samples were probed with monoclonal antibodies against β-actin as a loading control (Fig 2). Thus, we concluded that we successfully generated four Hdh-HET and four Hdh-KO MEF lines on similar genetic background. We reasoned that comparison of resulting MEF lines may reveal clues about normal function of Htt protein in cells.
<p>Figure 1</p>
Experimental procedure used to generate the Hdh-HET and Hdh-KO MEF cell lines
Experimental procedure used to generate the Hdh-HET and Hdh-KO MEF cell lines. The flow chart schematically describes the experimental paradigm used to generate Hdh-HET and Hdh-KO MEF cell lines. Please see text for details.
![]()
<p>Figure 2</p>
Western blotting analysis of Hdh-HET and Hdh-KO MEF lines
Western blotting analysis of Hdh-HET and Hdh-KO MEF lines. The whole cell lysates from 4 Hdh-HET and 4 Hdh-KO MEF cel lines were probed with anti-Htt monoclonal antibody. The same samples were also probed with anti-actin monoclonal antibody as a loading control. The ECL was used for detection.
![]()
Intracellular calcium signaling in Hdh-HET and Hdh-KO MEF cell lines
Previous studies have implicated impaired calcium signaling in the pathogenesis of HD 202122. The Htt directly binds to the inositol 1,4,5-trisposphate receptor (InsP3R), an intracellular Ca2+ release channel 2324. The expression of mutated Htt has been shown to affect the InsP3R activity 23 and mitochondrial Ca2+ signals and bioenergetics 252627282930. Since Htt and mutated Htt directly targets both ER and mitochondrial sites, it is possible that Htt may have some relevance for the physical and local Ca2+ coupling between ER and mitochondria 3132. To test this idea and to determine a role played by Htt in intracellular Ca2+ signaling, we performed a series of cytosolic and mitochondrial Ca2+ imaging experiments with generated MEF lines. Two Hdh-HET (HET1 and HET 5) and two Hdh-KO MEF cell lines (KO12 and KO27) (Fig 2) were selected for Ca2+ imaging studies.
In these experiments Hdh-HET and Hdh-KO MEF cell lines were challenged by ATP, an agonist of InsP3 signaling pathway in fibroblasts. Before stimulation with ATP the incubation medium was switched to a Ca2+ free buffer to prevent Ca2+ entry. The cytosolic [Ca2+]c and mitochondrial [Ca2+]m levels were monitored simultaneously as described in Methods. We found that pre-stimulation [Ca2+]c was higher in the Hdh-HET cells (HET1, 183 ± 11 nM (n = 42) and HET5 139 ± 9 nM (n = 33)) than that in the Hdh-KO cells (KO12, 102 ± 12 nM (n = 15) and KO27, 106 ± 6 nM (n = 48), p < 0.003). Stimulation with a suboptimal dose of ATP (2 μM) elicited large [Ca2+]c spikes in Hdh-HET cells, whereas only a small and slow [Ca2+]c rise was evoked in the Hdh-KO cells (Fig 3, Δ[Ca2+]c: HET1, 374 ± 24 nM; HET5, 386 ± 40 nM; KO12, 266 ± 47 nM and KO27, 284 ± 25 nM; p < 0.03). Many Hdh-KO cells did not show any [Ca2+]c rise in response to 2 μM ATP (not shown). Mobilization of the residual Ca2+ by optimal ATP (100 μM) evoked a relatively large [Ca2+]c elevation in the Hdh-KO cells (Fig 3). The ATP-induced [Ca2+]c spikes were closely followed by a [Ca2+]m elevation in the Hdh-HET cells, whereas the [Ca2+]m rise was modest in the Hdh-KO cells (Fig 3). Thus, the ATP-induced intracellular Ca2+ mobilization was suppressed and desensitized in the Hdh-KO cells.
<p>Figure 3</p>
Cytoplasmic and mitochondrial Ca2+ signals in Hdh-HET and Hdh-KO cells
Cytoplasmic and mitochondrial Ca2+ signals in Hdh-HET and Hdh-KO cells. MEF cells were transfected with mitochondrial matrix-targeted inverse pericam and were loaded with Fura-2/AM to monitor [Ca2+]m and [Ca2+]c, respectively. A. [Ca2+]c and [Ca2+]m signals in HET5 and KO27 cells. In the images, the inverse pericam fluorescence is shown in the gray scale; the blue overlay indicates the sites of the ATP-induced [Ca2+]m elevations (upper row of images). The Fura2 images are presented as green-red overlay where the [Ca2+]c elevations are indicated by a green to red shift (lower row). The graphs show the calibrated [Ca2+]c signal and the pericam fluorescence response (ΔFmito-pcam, decrease normalized to the initial fluorescence level) to the sequential stimulation by low (loATP, 2 μM) and high (hiATP, 100 μM) ATP in the single cells marked by the numbers. B. Mean [Ca2+]c and [Ca2+]m signals in Hdh-HET and Hdh-KO MEF cells. Traces show the mean of at least triplicate measurements for each cell line. The data are representative of four independent experiments.
![]()
Figure 3 has also showed that the [Ca2+]c signal was effectively propagated to the mitochondria in the Hdh-HET cells but less Ca2+ was transferred into the mitochondria in the Hdh-KO cells. To determine whether this resulted from the attenuated ER Ca2+ mobilization or from impaired ER-mitochondrial Ca2+ coupling, the [Ca2+]m rise was plotted against the [Ca2+]c increase for each cell., The relationship between [Ca2+]c elevations and the ensuing [Ca2+]m signals was similar in both the Hdh-HET and in the Hdh-KO cells (not shown). This result indicated that the local coupling between ER and mitochondria is maintained in the Hdh-KO cells.
The attenuated ATP-induced [Ca2+]c signal in the Hdh-KO cells could be due to reduced InsP3 generation, due to reduced sensitivity of InsP3R or due to depleted ER Ca2+ pool. To discriminate between these possibilities the InsP3-induced Ca2+ mobilization was quantified in suspensions of permeabilized MEF cells. The steady state [Ca2+]c was similar in both Hdh-HET and in Hdh-KO cells (Fig 4A, HET1, n = 12; HET5, n = 12; KO12, n = 13; KO27, n = 10 measurements). The Ca2+ pool size for both the ER and the ionophore-sensitive compartment was larger in the Hdh-KO than in the control cells (Fig 4A), whereas the uncoupler-sensitive mitochondrial compartment showed no difference (n = 3; not shown). Sequential application of a suboptimal and optimal InsP3 revealed lesser InsP3-sensitivity in the Hdh-KO cells than in the HET cells (Fig 4B, 4C, p < 0.01). However, the InsP3 sensitive fraction of the ER was approximately 75% in both control and Hdh-HET cells (Fig 4D). Furthermore, neither the passive Ca2+ buffering nor the mitochondrial Ca2+ uptake was altered in the cells lacking the Hdh (Fig 4E, 4F). Collectively, the data obtained in permeabilized cells suggest that the InsP3 sensitivity of the InsP3 receptor is attenuated in the Hdh-deficient cells, providing a mechanism to underlie the suppression of the InsP3-linked [Ca2+]c signaling in the Hdh-KO cells. The effect of Hdh on the InsP3 sensitivity is likely to be mediated by direct association between Htt and InsP3R 2324.
<p>Figure 4</p>
Ca2+ handling by Hdh-HET and Hdh-KO cells
Ca2+ handling by Hdh-HET and Hdh-KO cells. Fluorimetric measurements of cytoplasmic Ca2+ using Fura-2 (A-D) or FuraFF (E and F) in permeabilized Hdh-HET and Hdh-KO MEF cells were conducted as described in Materials and Methods. A. Steady state [Ca2+]c values and the size of the intracellular calcium pools in HET1, HET5, KO12 and KO27 cells. Prestimulation steady state [Ca2+]c obtained in presence of ATP 2 mM, creatine phosphate 5 mM, creatine phosphokinase 5 units/ml and succinate 2 mM. To obtain the size of InsP3, ER and total ionomycin-sensitive pools, 7.5 μM InsP3, 2 μM thapsigargin (Tg) and 10 μM ionomycin were added respectively. Mean ± SE of at least five independent experiments with multiple parallels (HET1: n = 12; HET5; n = 12; KO12, n = 13; KO27, n = 10). B. Analog traces showing the [Ca2+]c responses to suboptimal (250 nM) and optimal InsP3 (7.5 μM) and Tg for HET1, HET5, KO12 and KO27 cells. Changes were normalized to the size of the ER Ca2+ pool. C. Suboptimal InsP3 (250 nM)-induced [Ca2+]c increase normalized to total InsP3-sensitive (7.5 μM) pool. D. Maximal InsP3 (7.5 μM)-induced [Ca2+]c increase normalized to total Tg-sensitive (2 μM) ER pool. E. passive Ca2+ buffering assessed as the [Ca2+]c increase evoked by the addition of 10 μM CaCl2 in Tg and Ruthenium Red (3 μM)- pretreated MEF cells. F. Mitochondrial Ca2+ uptake caused by the addition of 40 μM CaCl2. Rate of uptake was calculated for initial 30 seconds after the challenge during which the rate was linear. Correction was made for the Ruthenium Red insensitive component. Data values show mean ± standard error of at least three independent experiments with multiple parallels. * indicates statistically significantly different p values (p < 0.03).
![]()
Microarray analysis of transcripts expressed in Hdh-KO and Hdh-HET MEFs
The results described in the previous section suggested that Htt plays a direct role in Ca2+ signaling by modulating InsP3R function. Many studies suggested that Htt also plays a major role in control of gene transcription 3334. To uncover potential gene expression changes we performed genome-wide transcription profiling of Hdh-HET and Hdh-KO MEF cells. Using the procedures described in Methods, we isolated total RNA from Hdh-HET MEF lines 1, 2 (in duplicate), 3, and 5 (in duplicate) and from Hdh-KO MEF lines 11, 12, 16 (in duplicate), 27 (in duplicate). The resulting 12 samples were provided to the UT Southwestern Microarray Core Facility (MCF) for genome-wide expression profiling using Sentrix Mouse-6 Expression Bead Chips (Illumina) (see Methods for details).
All 12 arrays produced highly reproducible and consistent gene expression data. The microarray results have been deposited in NCBIs Gene Expression Omnibus (GEO), and are accessible through GEO Series accession number GSE11139 35. Cluster analysis of the over-all gene expression data from the 12 samples demonstrated that all six Hdh-HET samples and all six Hdh-KO samples were clustered together, forming two clearly distinct groups (Fig 5). We also found that the duplicate samples (HET5.1 and HET5.2, HET2.1 and HET2.2, KO16.1 and KO16.2, and KO 27.1 and KO27.2) were most similar to each other when compared to other samples (Fig 5), as should be expected. Thus, we concluded that we obtained a high quality dataset of transcripts expressed in Hdh-HET and Hdh-KO MEF lines.
<p>Figure 5</p>
Cluster analysis of microarray data
Cluster analysis of microarray data. Using the Illumina BeadStudio 1.5 software package, each of 12 arrays was treated as an independent experiment and cluster analysis was performed on the microarray hybridization results.
![]()
In the next level of analysis, we combined all results obtained in six arrays with the Hdh-HET samples (HET group) and in six arrays with the Hdh-KO samples (KO group). Using Illumina BeadStudio software, we performed a statistical analysis to identify the differentially expressed genes between the Hdh-KO and Hdh-HET groups. We found that from 45992 probes existing on the Mouse-6 BeadChips arrays, 14,065 probes were present (with detection p-value < 0.01) in at least one of these two groups. Statistical analysis (t-test) has identified 821 transcripts that were significantly different between Hdh-HET and Hdh-KO groups (p < 0.05) (Fig 6). Among these 821 targets, 455 were up-regulated in the KO group and 366 were down-regulated (Fig 6). Thus, we concluded that inactivation of Hdh expression has a very significant effect on the transcriptional profile of MEF cells. The average signal intensities for each probe in Hdh-HET and Hdh-KO groups are included in the Excel format for the 821 differentially expressed targets (Additional files 1 and 2).
<p>Figure 6</p>
Altered gene expression between Hdh-HET and Hdh-KO groups
Altered gene expression between Hdh-HET and Hdh-KO groups. Cluster analysis of the 821 differentially expressed genes between Hdh-HET and Hdh-KO (p < 0.05). The genes were selected by comparing 6 HET samples with 6 KO samples using t-test with computing false discovery rate. The genes with p value <0.05 were selected and Hierarchical cluster analysis were performed using Cluster and TreeView http://rana.lbl.gov/EisenSoftware.htm. Each column represents a sample and each row represents a gene. The colorgram depicts high (red) and low (green) relative levels of gene expression in each sample.
![]()
<p>Additional file 1</p>
Supplementary Table 1. The list of annotated genes significantly affected in Hdh-KO MEF cells. The genes are classified into significant GO categories (classes) as explained in the text. The official gene symbol and protein name are shown. Also shown is average intensity of microarray signal for Hdh-HET and Hdh-KO samples and the calculated Diff score and p-values. Negative Diff scores correspond to genes underexpressed in Hdh-KO cells, positive Diff Scores correspond to genes overexpressed in Hdh-KO cells. The fold change is calculated by dividing average Hdh-HET and Hdh-KO signals. The transcripts which are not detected in Hdh-HET samples are shown as "up", the samples not detected in Hdh-KO samples are shown as "down". The Genebank accession numbers are also shown.
Click here for file
<p>Additional file 2</p>
Excell file with Supplementary Table 1. The microarray data for 821 differentially expressed probes are presented in the Excell file. The data listed for each probe are as follows: an Illumina beads target ID, an average signal from 6 chips with Hdh-HET samples, the probability of detection of a probe in Hdh-HET samples, an average signal from 6 chips with Hdh-KO samples, the probability of detection of a probe in Hdh-KO samples, the Diff Score obtained by comparing Hdh-HET and Hdh-HET samples, the P value calculated from the Diff score, an official gene symbol and definition (when available), the sequence of the probe, and the genebank accession number (when available). The results are sorted by the Diff score with negative Diff scores corresponding to the probes downregulated in Hdh-KO cells (366 probes) and positive Diff scores corresponding to the probes upregulated in Hdh-KO cells (455 probes).
Click here for file
Analysis and annotation of microarray data
In the annotation of microarray data, we focused on 279 unique genes whose expression differed by at least 2-fold between Hdh-HET and Hdh-KO groups. From these genes 173 genes (104 of which are annotated) were up-regulated and 106 genes (65 of which were annotated) were down-regulated in Hdh-KO cells when compared to Hdh-HET cells. In order to extract biological information from these genes, we performed annotation of the array results using the Ingenuity Pathway Analysis platform 36 and the GoStat software 37. We found that several significant GO categories can be pulled from the data: developmental process (35), nervous system development and function (19), lipid metabolism (18), glucosamine metabolic process (2), transcription regulator activity (5), cellular_component:plasma membrane (16), regulation of cellular process (29), endocytosis (2), mitochondrion (3), extracellular matrix (2), cytoskeleton (5), others (33), indicating the possible disruption of several functional pathways in the absence of Htt (Additional files 1 and 2). We have not observed significant changes in gene expression levels of most proteins involved in Ca2+ signaling pathways (Additional files 1 and 2), indicating that Ca2+ signaling changes observed in our functional experiments (Figs 2, 3, 4) are likely to be due to post-translational effects, such as for example changes in InsP3R gating properties.
Confirmation of gene expression with real-time PCR
In order to confirm our microarray results, we performed quantitative real-time PCR (qPCR) analysis for several of the candidate genes. For these experiments we choose three genes that were significantly down-regulated in Hdh-KO MEFs (Hdh, Sox-2, Tcf2) and three genes that were significantly up-regulated in Hdh-KO MEFs (Cart1, Esm1, Pitx2) (Additional files 1 and 2). We observed a good correlation between averaged microarray results and qPCR data for all six genes evaluated (Table 1). Moreover, we observed a good correlation between microarray and qPCR data for results obtained with cDNA samples from individual Hdh-HET and Hdh-KO MEF cell lines (Fig 7).
<p>Figure 7</p>
Correlation between microarray and qPCR results
Correlation between microarray and qPCR results. The results of the microarray and qPCR analyses are compared for 3 genes downregulated (Hdh, Tcf2, Sox2) and 3 genes upregulated (Cart1, Esm1, Pitx2) in Hdh-KO cells. The microarray and qPCR results are shown for RNA extracted from 6 Hdh-KO samples (blue) and 6 Hdh-HET samples (red).
![]()
<p>Table 1</p>
Comparison of average microarray and qPCR results for 6 selected genes. 3 genes downregulated in Hdh-KO cells are shown in blue, 3 genes upregulated in Hdh-KO cells are shown in red. The average signals obatined in Microarray and qPCR experiments are shown for each gene. Also shown are primers used for qPCR experiments and the fold change calculated from microarry and qPCR data by dividing average Hdh-KO signal to average Hdh-HET signal.
Gene
GeneBank Acc #
Microrray
Q-PCR
HET
KO
Fold
Primers (FP/RP)
HET
KO
Fold
Hdh
NM_010414
303.7
-0.9
DOWN
CGTGAGCATCTGCCAACATT
CACACCGAGGATCAGGAGAGT
404.3
2.0
-202.2
Sox-2
NM_011443
233.7
-5.6
DOWN
GATGCACAACTCGGAGATCAG
GCTTCTCGGTCTCGGACAA
16615.9
18.4
-903.0
Tcf2
NM_009330
255.6
7.3
-35.0
CACCTGACAGTAAAATGCAGATCA
GAGTGTGTGCTGTGGATATTCGT
4513.7
2.6
-1736.0
Cart1
NM_172553
-9.3
286.9
UP
CCCTGAGAACGGAGCTCACT
TTGGCCCTTCGATTTTGAA
12.5
45701.8
3656.1
Esm1
NM_023612
-35.6
498.6
UP
GCGAGGAGGATGATTTTGGT
GCATTCCATCCCGAAGGT
19.7
2966.9
150.6
Pitx2
NM_011098
124.8
1811.8
14.5
CCGAGTCCGGGTTTGG
TTGCGTTCCCGCTTTCTC
21.3
368.2
17.3
Comparison with gene expression profiling of Hdh-null ES cells
When our paper was in preparation, another group independently reported genome-wide expression profiling of Hdh-null embryonic stem (ES) cells 38. It is of interest to compare our findings with Hdh-KO MEF cells with results obtained for Hdh-null ES cells. Strehlow at al (2007) reported that expression of 16 known transcripts was significantly affected in Hdh-null ES cells when compared to wild type ES cells. The affected transcripts were divided into several classes of interest: protein degradation (5), extracellular matrix (4), cell division (4), and patterning/development (3) 38. To compare these data with our results, we attempted to locate the genes highlighted by the study by Strehlow at al (2007) among the genes which expression differs at p < 0.05 between Hdh-KO and Hdh-HET MEF cells in our experiments. We determined that two of the three genes in the patterning/development category (Otx2 and Pem) do not present in MEF cell lines, Leftb/Lefty is present in MEF cell lines, but did not show a significant difference between HET and KO MEFs. All five genes in the protein degradation category are present in MEF cell lines but did not show significant difference in our experiments. We found that Adam23 is not present in MEF cell lines. B3galt6, col4a1 and clo4a2 are present in MEF cells, but did not show significant difference. We further found that one out of four genes in cell division category (Ccdc5) is also significantly affected in our experiments. Ccdc5 is up-regulated in both Hdh-null ES cells and in Hdh-KO MEF cells. Plk1 is also strongly upregulated (8.7-fold) in Hdh-null ES cells 38 but did not show significant difference in Hdh-KO MEF cells in our experiments. Aurkc is not present in MEF cells, while Pttg1 is present but did not show significant difference.
Strehlow at al (2007) induced neuronal differentiation of Hdh-null ES cells by application of retinoic acid and performed microarray analysis of the in vitro differentiated neurons at 6, 8, and 10 days post-differentiation 38. The expression of transcripts in the number of categories was different between Hdh-null and wild type in vitro differentiated neurons 38. Changes in only a few genes have been consistently observed in the two studies (Table 2).
<p>Table 2</p>
Comparison of gene expression changes in Hdh-KO MEF and differentiated Hdh-null ES cells. The genes are classified into functional categories (groups) based on GO classification as explained by Strehlow at al (2007). The number of genes affected in differentiated Hdh-null ES cells is shown. The number of these genes present among the genes affected in our experiments with Hdh-KO MEF cells is also shown. The official gene symbols, accession numbers, Diff scores and fold changes are shown for all overlapping genes. The positive Diff score corresponds to the genes upregulated in Hdh-KO MEG cells, the negative Diff score corresponds to the genes downregulated in Hdh-KO MEF cells. The genes which expression is affected in the opposite direction when Hdh-null ES and Hdh-KO MEF data are compared are shown in italic.
Gene groups
Gene # HMG paper
overlap in SupTable 1
Overlap gene symbol
Accession
Diff_Score
change fold
extracellur space
99
1
ilvbl
NM_173751.3
0.037888
2.037322394
extracellular matrix/components conferring tensile strength/cell adhesion
38
2
Emp3 (Ymp)
NM_010129.1
85.9198
2.395867383
recepto binding/hormone activity
27
0
Epb4.1l3
NM_013813.1
23.5427
1.63780359
lysosome activity/protein degradation
17
0
enzyme inhibitor activity
15
0
growth, proliferation and differentiation
15
1
Emp3 (Ymp)
NM_010129.1
85.9198
2.395867383
apoptosis
10
1
Txnl1
NM_016792.2
0.003063
2.178257007
axonogenesis
10
0
synaptic activity
14
0
wnt signaling
6
0
metal ion transport
8
1
Crip1
NM_007763
0.049374
5.609568813
transport
4
0
retinol/retinal/retinoid/isoprenoid binding
5
0
Discussion
Despite the importance of understanding the normal function of Htt for both basic biology and for HD, its function remains largely unknown 934. The generation of Hdh-null MEF cell lines described in our study provides a new and unbiased approach to search for novel Htt functions.
The Htt directly binds to the InsP3R, an intracellular Ca2+ release channel 2324. The expression of mutated Htt has been shown to affect the InsP3R activity 23 and mitochondrial Ca2+ signals and bioenergetics 252627282930. Thus, in the first series of experiments we evaluated InsP3R-mediated Ca2+ cytosolic and mitochondrial Ca2+ signals in Hdh-null MEF cells. As a result of these experiments we found that InsP3R sensitivity to stimulation by InsP3 was reduced in the absence of Hdh (Figs 3 and 4). We further found that Htt appears to be dispensable for ER-mitochondrial Ca2+ coupling (data not shown). Thus the altered InsP3R-induced cytoplasmic and mitochondrial calcium signaling in the Hdh-null MEF cells may result from the lack of Hdh by itself and does not necessarily require a secondary change in gene regulation.
Interestingly, a large number of Ca2+-related genes, such as CACNA2D3 (calcium channel, voltage-dependent, alpha2/delta subunit 3), ITPR1 (inositol 1,4,5-trisphosphate receptor 1), HOMER1 (homer homolog 1), ATP2A2 (ATPase, Ca2+ transporting, cardiac muscle, slow twitch 2), DRD2 (dopamine receptor 2), PRKCB1 (protein kinase C beta 1), PDE1B (phosphodiesterase 1B, Ca2+-calmodulin dependent), ATP2B2 (ATPase, Ca2+ transporting, plasma membrane 2), CAMK2B (calcium/calmodulin-dependent protein kinase II, beta), PLCB1 (phospholipase C, beta 1), RGS4 (regulator of G-protein signaling 4), and CAMK2A (calcium/calmodulin-dependent protein kinase II alpha), have been recently reported to be consistently and significantly downregulated in a striatal region of symptomatic human HD patients and aging HD mouse models 39. These results are in agreement with the "Ca2+ hypothesis of HD" 22 and with a direct role of Htt in intracellular Ca2+ signaling supported by our experiments.
Many studies suggested that Htt also plays a major role in control of gene transcription 3334. To search for changes in gene transcription resulting from deletion of Htt gene, we performed a genome-wide comparison of transcription profiles in MEF cells expressing a single copy of Hdh (Hdh-HET cells) and in MEF cells which lack Hdh expression (Hdh-KO cells). To minimize sources of variability, the Hdh-HET and Hdh-KO MEF cells were generated in parallel experiments and on identical genetic background. From our annotation analysis, we found that a large group of affected genes play a role in embryonic development (Additional file 1). This result was not unexpected because Htt is essential for embryonic development, and complete inactivation of Htt expression in knock-out mice causes early embryonic lethality 101112. The functions of these genes may provide additional clues about the mechanism responsible for embryonic lethality in Hdh knockout mice, for example there are some similar phenotypic manifestations between Hdh nullizygous embryos and knockout mutants lacking fibroblast growth factor receptor1(fgfr1) 12. Interestingly, we found fgfrl1 message is downregulated approximately 2-fold in Hdh-KO MEF cells when compared to Hdh-HET cells (Additional files 1 and 2).
After gastrulation, Htt is important for neurogenesis – mice carrying a <50% dose of wild-type Htt display profound malformations of the cortex and striatum 13. The neuronal inactivation of Htt during mid- to late gestation, for example, leads to neurological abnormalities and progressive degeneration (apoptotic cells in the hippocampus, cortex and striatum, and a lack of axons) 18. Our analysis revealed a number of genes involved in nervous system development and function which were affected in Hdh-KO MEF cell lines (Additional file 1), and this list should also provide useful information to guide further studies of Htt's normal function in the nervous system. For example, Sox-2 expression was absent from Hdh-KO MEFs. Sox (Sry-related HMG box) genes encode transcription factors regulating crucial developmental decisions in different systems. Sox2 is expressed in, and is essential for, totipotent inner cell mass stem cells and other early multipotent cell lineages, and its ablation causes early embryonic lethality 40. In many different species, Sox2 is a marker of the nervous system from the beginning of its development, it maintains a stem-cell like state and actively inhibits neuronal differentiation, Sox2 deficiency causes neurodegeneration and impaired neurogenesis in the adult mouse brain. Does the absence of Sox-2 play some role in the early embryonic lethality and neurodegeneration in Hdh knock-out mice and conditional knock-out mice respectively? Further studies are required to answer these questions. Sox-11 is another sox family gene which changes dramatically in Hdh-KO MEF cell lines (increases about 10-fold, see Additional files 1 and 2). The widespread expression of sox-11 in both the central and peripheral nervous system suggests that sox-11 plays a general role in neuronal development, and its changes in Hdh-KO cells merit further investigation.
Another group of genes whose expression was significantly affected in Hdh-KO MEF cells are the genes related to lipid metabolism. It has been reported in another microarray analysis using clonal striata-derived cells, that genes involved in lipid metabolism were affected after expressing different N-terminal 548-amino-acid Htt fragments 41. Moreover, recent biochemical data indicated that Htt binds to caveolin and plays a direct role in cholesterol metabolism 42. All these data suggested that Htt plays an important role in lipid metabolism, which may be affected by HD mutation. Indeed, RXRG (Retinoic acid receptor RXR-gamma) and RBP4 (retinol binding protein 4) are consistently downregulated in a striatal region of symptomatic human HD patients and aging HD mouse models 39.
From our analysis we also found calcium channel voltage-dependent alpha2/delta subunit 1 (Cacna2d1) was down-regulated about 2 fold (Additional files 1 and 2), interestingly the same CACNA2D1 protein has been recently identified as novel Htt-binding partner in unbiased mass-spectroscopy screen 24. A closely related alpha2/delta subunit 3 (Cacna2d3) was reported on the 3rd place on the list of the genes which are consistently and significantly downregulated in a striatal region of symptomatic human HD patients and aging HD mouse models 39. As discussed above, these results indicate that Htt may play a role in regulation of Ca2+ channel activity and Ca2+ signaling in cells, consistent with Ca2+ hypothesis of HD 2243.
Methods
Generation of Hdh-HET and Hdh-KO MEF cell lines
Generation and characterization of the Hdh +/- and Hdh-floxed mice have been described previously 1218. E13.5 embryos obtained from a cross between the Hdh-floxed and Hdh +/- mice were first eviscerated and decapitated, and then the carcasses were finely minced using scissors. The tissue obtained from each embryo was digested with 0.25% Trypsin-EDTA at 37°C for 10 min, washed once with 10% FBS in DMEM, and the cell suspensions from each embryo were plated separately in 10% FBS-DMEM culture medium in order to obtain cultures of primary fibroblasts. Following plating of the cells, the genotype of each embryo was determined by PCR 1218. After two days in culture, the primary fibroblasts from each genotype (Hdh floxed/+ and Hdh floxed/-) were pooled together. The pooled cells were then plated on six-well tissue culture plates, grown to 60–80% confluence and transfected with SV40-Large T-antigen plasmid in pcDNA3-Zeo vector (linearized with PvuI) using the Fugene-6 lipofection reagent (Roche). Transfected cells were cultured for four-six weeks until immortalized Hdh floxed/+ and Hdh floxed/- mouse embryonic fibroblasts (MEFs) were obtained. Hdh floxed/+ and Hdh floxed/- immortalized MEFs were then infected with Lenti-NLS-GFP-Cre virus 19 encoding a nuclear-targeted GFP-Cre fusion protein. Expression of NLS-GFP-Cre in infected cells catalyzed recombination at the Hdh(flox) loxP sites leading to excision of the promoter and exon 1 sequences of Hdh(flox) allele 18. Following infection with Lenti-NLS-GFP-Cre, the Hdh floxed/+ and Hdh floxed/- MEFs were plated in 10% FBS-DMEM culture medium in 96 well plates at an average density of one cell/well for clonal selection. Successful recombination with NLS-GFP-Cre converts the Hdh floxed/+ MEFs to Hdh +/- (Hdh-HET) MEFs and the Hdh floxed/- MEFs to Hdh-/-(Hdh-KO) MEFs. After four weeks in culture, several potential Hdh-HET and Hdh-KO MEF lines were selected, expanded and analyzed by Western blotting with monoclonal antibodies against Hdh (Chemicon MAB2166, 1:500) and monoclonal antibodies against actin (Chemicon MAB1501, 1:2000).
Cytosolic and mitochondrial Ca2+ imaging
For measurements of mitochondrial matrix [Ca2+] ([Ca2+]m), the cells were transfected with a mitochondrial matrix targeted inverse pericam construct 44 by electroporation 48–72 h prior to the imaging experiment. Before use, the cells were preincubated in an extracellular medium (2% BSA/ECM) consisting of 121 mM NaCl, 5 mM NaHCO3, 10 mM Na-HEPES, 4.7 mM KCl, 1.2 mM KH2PO4, 1.2 mM MgSO4, 2 mM CaCl2, 10 mM glucose and 2% bovine serum albumin (BSA), pH 7.4. To monitor [Ca2+]c cells were loaded with 2.5 μM Fura2/AM for 20–30 min in the presence of 200 μM sulfinpyrazone and 0.003% (w/v) pluronic acid at room temperature. Before start of the measurement the buffer was replaced by a Ca2+ free 0.25%BSA/ECM ([Ca2+] <1 μM).
Coverslips were mounted on the thermo stated stage (35°C) of an Olympus IX70 inverted microscope fitted with a 40× (UApo, NA 1.35) oil immersion objective. Fluorescence images were collected using a cooled CCD camera (Pluto, Pixelvision). Excitation was rapidly switched among 340 and 380 nm for fura2 and 495 nm for pericam and a 510 nm longpass dichroic mirror and a 520 nm longpass emission filter were used.
For evaluation of cytoplasmic [Ca2+] ([Ca2+]c), Fura2 fluorescence was calculated for the total area of individual cells. [Ca2+]c was calibrated in terms of nM using in vitro dye calibration. For evaluation of [Ca2+]m the pericam-mt signal was masked. Recordings obtained from all transfected cells on the field (2–10 cells) were averaged for comparison in each experiment. Experiments were carried out with at least four different cell preparations. Significance of differences from the relevant controls was calculated by ANOVA.
Measurements of cytosolic Ca2+ in permeabilized cells
The cells (2 mg protein/1.5 ml) were permeabilized in an intracellular medium (KCl 120 mM, NaCl 10 mM, KH2PO4 1 mM, Tris-HEPES 20 mM, and antipain, leupeptin and pepstatin 1 μg/ml each at pH 7.2) supplemented with 40 μg/ml digitonin and with fura2/FA (1.5–3 μM) or furaFF/FA (0.5 μM) and TMRE 2 μM in a stirred thermo stated cuvette at 35°C. All the measurements were carried out in the presence of succinate 2 mM, 2 mM MgATP and ATP regenerating system composed of 5 mM phosphocreatine, 5 U/ml creatine kinase. Fura2/FA or FuraFF/FA was added to monitor [Ca2+] in the intracellular medium that exchanges readily with the cytosolic space and so [Ca2+]fura2 was abbreviated as [Ca2+]c. Fluorescence was monitored in a fluorometer (Delta-RAM, PTI) using 340 nm, 380 nm excitation and 500 nm emission for fura2FF and 540 nm excitation and 580 nm emission for TMRE. Calibration of the fura fluorescence was carried out at the end of each measurement as described previously 45. Experiments were with at least five different cell preparations in multiple parallels. Significance of differences from the relevant controls was calculated by ANOVA.
Microarray analysis
Total RNA was isolated from fibroblast cultures using the TRIZOL reagent according to manufacturer's instructions (Invitrogen). Briefly, the MEF cells were grown to 60–80% confluence in T25 tissue culture flask, the culture medium was aspired and 1 ml of TRIZOL reagent was added to each flask. The cells were incubated with TRIZOL at room temperature for 5 min. The resulting lysates were collected from each flask, mixed with 0.2 ml of chloroform and centrifuged at 12,000 × g for 15 min at 4°C. The supernatants were collected, mixed with an equal volume of 70% ethanol at room temperature and immediately transferred to RNAeasy mini spin columns for RNA purification according to the manufacturer's (Qiagen) instructions. The final RNA samples were eluted from the RNeasy mini spin columns in 30 μl of DEPC-treated water. Using the procedures described above, we isolated total RNA from Hdh-HET MEF lines 1, 2 (in duplicate), 3, and 5 (in duplicate), and from Hdh-KO MEF lines 11, 12, 16 (in duplicate), 27 (in duplicate). The resulting 12 samples were submitted to the UT Southwestern Microarray Core Facility for microarray analysis. Biotinylated cRNA was prepared using the Illumina RNA Amplification Kit (Ambion, Inc., Austin, TX) according to the manufacturer's directions starting with ~200 ng total RNA. Samples were purified using the RNeasy kit (Qiagen, Valencia, CA). Hybridization to the Sentrix Mouse-6 Expression BeadChip (Illumina, Inc., San Diego, CA), washing and scanning were performed according to the Illumina BeadStation 500× manual (revision C). Two BeadChips were used, each one containing 6 arrays. For each chip, three HET and three KO samples were analyzed to minimize the effects of chip-to-chip variability. Arrays were scanned with an Illumina Bead array Reader confocal scanner and the data was analyzed using Illumina's BeadStudio software(Version 3). The raw data was background subtracted and normalized using "cubic spline" method in the software. The detection p values were computed using a dynamically constructed normal model based on the intensities of 700 negative controls. For differential analysis, the six arrays in HET group were compared with the six arrays in KO group using t-test with computing false discovery rate algorithm. The genes with p value <0.05 were considered differentially expressed and subject for further analysis.
Quantitative Real Time PCR
Quantitative real time (qRT)-PCR was performed using an Applied Biosystems Prism 7900HT sequence detection system using SYBR green chemistry. Briefly, total RNA was treated with DNase I (RNase-free, Roche Molecular Biochemicals), and reverse-transcribed with random hexamers using SuperScript II reverse transcriptase (Invitrogen) to generate cDNA as previously described 46. Primers were designed using Primer Express Software (PerkinElmer Life Sciences) and validated by analysis of template titration and dissociation curves. Each qRT-PCR contained (final volume of 10 μl) 25 ng of reverse-transcribed RNA, each primer at 150 nM, and 5 μl of 2× SYBR Green PCR Master Mix (Applied Biosystems), and each sample was analyzed in triplicate. Results were evaluated by the comparative CT method (User Bulletin No. 2, PerkinElmer Life Sciences) using cyclophilin as the invariant control gene.