SUPPORTING INFORMATION

Tissue-specificity of dystrophin-actin interactions: Isoform-specific thermodynamic

stability and actin binding function of tandem calponin-homology domains
Vaibhav Upadhyay, Swati Bandi, Sudipta Panja, Laura Saba, and Krishna M.G. Mallela*
Department of Pharmaceutical Sciences, Skaggs School of Pharmacy and Pharmaceutical
Sciences, University of Colorado Anschutz Medical Campus, Aurora, CO 80045.

Table S1: Thermodynamic parameters of tandem CH domains obtained from individual fitting
of urea denaturation melts using CD and fluorescence spectroscopy.
Isoform Urea denaturation Urea denaturation
(CD) (Fluorescence)
ΔG0unf Cm m-value ΔG0unf Cm m-value
(kcal/mol) (M [urea]) (kcal/mol/M (kcal/mol) (M [urea]) (kcal/mol/M
[urea]) [urea])
Dys-M 11.90 ± 0.34 5.58 ± 0.01 -2.13 ± 0.06 11.88 ± 0.55 5.54 ± 0.02 -2.14 ± 0.10
Dys-P 6.30 ± 0.24 5.35 ± 0.03 -1.18 ± 0.05 6.38 ± 0.52 5.19 ± 0.02 -1.23 ± 0.10
Dys-B 6.90 ± 0.17 5.21 ± 0.02 -1.32 ± 0.03 6.83 ± 0.47 5.12 ± 0.04 -1.33 ± 0.09

Table S2: Kd values obtained for interaction of dystrophin tandem CH domain isoforms with
skeletal muscle actin.
Isoform Kd (µM)
Dys-M 130 ± 7
Dys-P 87 ± 4
Dys-B 109 ± 3

S1
Figure S1: Actin binding of tandem CH domains of dystrophin isoforms. (A-C) SDS-PAGE of
the pellets from high-speed centrifugation performed at a fixed concentration of F-actin (7 µM)
and with varying concentrations of tandem CH domains of Dys-M, Dys-P, or Dys-B. (D) Actin
binding curves obtained from the band intensities on SDS-PAGE shown in panels A-C, after
correcting for differential staining of the dye to proteins.

S2
3-state analysis of urea denaturation melts of dystrophin tandem CH domain isoforms
Since tandem CH domains unfold through an intermediate state with only CH1 domain
unfolded,1-2 denaturant melts of the three isoforms should in principle be analyzed in terms of a
3-state equilibrium,
𝑁𝑁 ⇌ 𝐼𝐼 ⇌ 𝑈𝑈
where N represents the native state with both CH1 and CH2 folded, I represents the intermediate
state with CH1 unfolded and CH2 folded, and U represents the unfolded state with both CH1 and
CH2 unfolded. In the case of such 3-state equilibrium, the equation that describes the denaturant
melt becomes,3-4

𝑆𝑆𝐷𝐷
0 0 0
∆𝐺𝐺𝑁𝑁𝑁𝑁 +𝑚𝑚𝑁𝑁𝑁𝑁 [𝐷𝐷] ∆𝐺𝐺𝑁𝑁𝑁𝑁 +𝑚𝑚𝑁𝑁𝑁𝑁 [𝐷𝐷] ∆𝐺𝐺𝐼𝐼𝐼𝐼 +𝑚𝑚𝐼𝐼𝐼𝐼 [𝐷𝐷]
(𝑆𝑆𝑁𝑁 + 𝑚𝑚𝑁𝑁 [𝐷𝐷]) + (𝑆𝑆𝐼𝐼 + 𝑚𝑚𝐼𝐼 [𝐷𝐷])𝑒𝑒 − 𝑅𝑅𝑅𝑅 + (𝑆𝑆𝑈𝑈 + 𝑚𝑚𝑈𝑈 [𝐷𝐷])𝑒𝑒 − 𝑅𝑅𝑅𝑅 𝑒𝑒 − 𝑅𝑅𝑅𝑅
= 0 +𝑚𝑚 [𝐷𝐷]
∆𝐺𝐺𝑁𝑁𝑁𝑁 𝑁𝑁𝑁𝑁
0 +𝑚𝑚 [𝐷𝐷]
∆𝐺𝐺𝑁𝑁𝑁𝑁 𝑁𝑁𝑁𝑁 ∆𝐺𝐺 0 +𝑚𝑚𝐼𝐼𝐼𝐼 [𝐷𝐷]
− − − 𝐼𝐼𝐼𝐼
1 + 𝑒𝑒 𝑅𝑅𝑅𝑅 + 𝑒𝑒 𝑅𝑅𝑅𝑅 𝑒𝑒 𝑅𝑅𝑅𝑅

(Eq. S1)
where SD is the measured signal as a function of the denaturant concentration D; SN, SI, and SU are
the intrinsic signals corresponding to the native, intermediate, and unfolded states in the absence
of denaturant; mN, mI, and mU are the slopes of the linear dependence of SN, SI, and SU on
denaturant concentration; ∆G0NI and ∆G0IU correspond to the Gibbs free energies of the first (N ⇌
I) and second (I ⇌ U) transitions in the absence of denaturant, and mNI and mIU are the slopes of
linear dependence of ∆GNI and ∆GIU with denaturant concentration.
Above Eq. S1 indicates that 3-state fit of a denaturant melt involves ten fit parameters,
specifically, SN, mN, SI, mI, SU, mU, ∆G0NI, mNI, ∆G0IU, and mIU. Unless the protein shows a clear
double sigmoidal denaturant melt with two well-separated transitions corresponding to the
unfolding of CH1 and CH2 domains or without knowing the spectral properties of the intermediate
state and their denaturant dependence, it is not possible to fit the denaturant melts shown in Figs.
3A and 3B to a unique set of parameters with much less standard errors of estimates on fit
parameters. Therefore, to minimize the number of fit parameters, we need to make reasonable
assumptions. Since the intermediate consists of unfolded CH1 domain and folded CH2 domain,
CD signal at 222 nm, which measures the α-helical secondary structure, of the intermediate and
its denaturant dependence is assumed to be half that of the native and unfolded states (SI =
(SN+SU)/2; mI = (mN+mU)/2). This assumption may not be valid in the case of denaturant melts
measured by fluorescence, because the change in fluorescence signal may not mirror the equivalent
change in the protein’s structure, since fluorescence signals depend primarily on the solvent
environment and the nearby chemical groups that quench the fluorescence.5-6 In addition, muscle
isoform contains two additional tryptophans compared to brain and Purkinje isoforms (Fig. 1B);
hence fluorescence signal of the intermediate may not be half of the corresponding native and
unfolded states. Therefore, we analyzed only the denaturant melts obtained by CD to the above 3-
state equation (Eq. S1). Since the difference between the three isoforms is only in the N-terminal
flanking region of the CH1 domain, we also assumed that the stability of CH2 domain with

S3
unfolded CH1 is the same across the three isoforms. Figure S2 shows the fit and Table S3 lists the
fit parameters.

1.0
Normalized CD @ 222 nm

0.8

0.6

0.4

0.2
Dys-M
Dys-P
Dys-B
0.0

0 2 4 6 8
[Urea] (M)
Figure S2: Fitting of CD denaturant melts to 3-state equilibrium equation S1. Solid lines represent
the fit curves with constraints listed in Table S3.

Table S3: Thermodynamic stability parameters obtained from the global analysis of CD
denaturant melts.
Fit constraints Tandem ∆G0NI = mNI = mCH1 + ∆G0IU = mIU = mCH2
CH domain ∆G0CH1 + mCH1-CH2 ∆G0CH2 (kcal/mol/M
isoform ∆G0CH1-CH2 (kcal/mol/M (kcal/mol) [urea])

(kcal/mol) [urea])

SI = (SN+SU)/2 Dys-M 11.55 ± 1.10 -2.01 ± 0.23 2.79 ± 1.44 -0.51 ± 0.34
mI = (mN+mU)/2
Dys-P 5.73 ± 1.19 -1.04 ± 0.23 2.79 ± 1.44 -0.51 ± 0.34
Dys-B 6.51 ± 0.93 -1.22 ± 0.18 2.79 ± 1.44 -0.51 ± 0.34

S4
Although the data fit well to the 3-state model with a reasonable constraint about intermediate
CD signal and its denaturant dependence, the absence of a well-separated second transition
resulted in higher standard errors of estimates on fit values of ∆G0IU and mIU.
In the above fit (Fig. S2 and Table S3), ∆G0NI corresponds to the unfolding of the CH1
domain, whereas ∆G0IU corresponds to the unfolding of the CH2 domain. For those tandem CH
domains such as that of dystrophin which exists in a closed conformation with significant
interactions between folded CH1 and CH2 domains, when CH1 unfolds, it also results in the
breaking of the inter-CH-domain interactions. Therefore, the ∆G0NI becomes the sum of intrinsic
stability of CH1 domain in the presence of CH2 (∆G0CH1) and the strength of CH1 interactions
with CH2 (∆G0CH1-CH2), as stated in Table S3. Since CH1 domain determines the actin binding
affinity of tandem CH domains and inter-CH-domain interactions determine the thermodynamic
stability of tandem CH domains rather than their actin-binding affinity,7-9 we need to know the
intrinsic stability of CH1 in the full-length dystrophin tandem CH domain (∆G0CH1) that is
distinct from that of inter-CH-domain interactions (∆G0CH1-CH2) to establish a quantitative
correlation between the thermodynamic stability and actin-binding affinity of dystrophin tandem
CH domain isoforms, which is not possible with the current data analysis. We have shown
previously that such quantitative relationship exists in the case of utrophin tandem CH domain,
which exists in an open conformation with minimal inter-CH-domain interactions (∆G0NI ≈
∆G0CH1; ∆G0CH1-CH2 ≈ 0); ∆∆G0CH1 is of the same magnitude as that of ∆∆G0 of actin binding.3

S5
REFERENCES
1. Bandi, S.; Singh, S. M.; Mallela, K. M. G., The C‑terminal domain of the utrophin
tandem calponin-homology domain appears to be thermodynamically and kinetically more stable
than the full-length protein. Biochemistry 2014, 53 (14), 2209-11.
2. Singh, S. M.; Molas, J. F.; Kongari, N.; Bandi, S.; Armstrong, G. S.; Winder, S. J.;
Mallela, K. M. G., Thermodynamic stability, unfolding kinetics, and aggregation of the N-
terminal actin-binding domains of utrophin and dystrophin. Proteins: Struct. Funct. Bioinform.
2012, 80 (5), 1377-92.
3. Singh, S. M.; Bandi, S.; Mallela, K. M. G., The N-Terminal Flanking Region Modulates
the Actin Binding Affinity of the Utrophin Tandem Calponin-Homology Domain. Biochemistry
2017, 56 (20), 2627-2636.
4. Bilsel, O.; Zitzewitz, J. A.; Bowers, K. E.; Matthews, C. R., Folding mechanism of the α-
subunit of tryptophan synthase, an α/β barrel protein: global analysis highlights the
interconversion of multiple native, intermediate, and unfolded forms through parallel channels.
Biochemistry 1999, 38, 1018-1029.
5. Chen, Y.; Barkley, M. D., Toward understanding tryptophan fluorescence in proteins.
Biochemistry 1998, 37 (28), 9976-82.
6. Royer, C. A., Probing protein folding and conformational transitions with fluorescence.
Chem. Rev. 2006, 106, 1769-1784.
7. Singh, S. M.; Bandi, S.; Winder, S. J.; Mallela, K. M. G., The actin binding affinity of the
utrophin tandem calponin-homology domain is primarily determined by its N‑terminal domain.
Biochemistry 2014, 53 (11), 1801-9.
8. Bandi, S.; Singh, S. M.; Mallela, K. M. G., Interdomain linker determines primarily the
structural stability of dystrophin and utrophin tandem calponin-homology domains rather than
their actin-binding affinity. Biochemistry 2015, 54 (35), 5480-8.
9. Singh, S. M.; Bandi, S.; Mallela, K. M. G., The N- and C-terminal domains differentially
contribute to the structure and function of dystrophin and utrophin tandem calponin-homology
domains. Biochemistry 2015, 54 (46), 6942-50.

S6