Identification of Amino Acids that are Critical for Structural Stability and Functionality within the Heterodimerization (HD) Domain of Notch Proteins

1
Identification of Amino Acids that are Critical
for Structural Stability and Functionality within
the Heterodimerization (HD) Domain of Notch
Proteins Marina Pellón Consunji and Lucien
Celine Montenegro Advisor Dr. Didem
Vardar-Ulu Wellesley College, Massachusetts
Table 1. For every amino acid position in hN2 HD
WT, report of prevalence of physicochemical
properties. 126 protein sequences found to be
common to the BLAST results of hN1, hN2, and hN3
were multiple sequence aligned based on BLAST
results to hN2 HD. Each of the four columns for
every amino acid position report the prevalence
of one of the four physicochemical properties
studied (PS, NG, PL, NP). White lt 25 yellow
25-50 orange 50-75 red gt75. For specific
amino acid groupings, see Python under
Experimental Methods. Marked with a red arrow are
L1573P and L1625P, positions that have been
mutated in this study. 7,8
Results
Introduction
The integral role Notch proteins play in the
normal development and tissue homeostasis of
metazoans is orchestrated through the tightly
regulated proteolytic processing of their
Negative Regulatory Region (NRR) (Figure 1).
Ligand binding at an extracellular region
activates Notch by facilitating a proteolytic
cleavage at the heterodimerization (HD) domain
within the NRR. This cleavage is a prerequisite
for the subsequent intramembrane cleavage that
permits the translocation of the intracellular
region of Notch to the nucleus to activate
transcription. In this work we present the
algorithms we have developed to evaluate the
relative importance of specific amino acids for
the structural stability and functionality of the
human Notch 2 (hN2) HD domain. These include
predictive models derived from bioinformatic
analysis as well as experimental data from
circular dichroism (CD), C18 reverse phase high
performance liquid chromatography (C18 HPLC), and
mass spectrometry (MS) studies that compare the
wild-type HD domain with L1573P and L1625P, two
mutants identified in leukemia patients.1,2 Our
results provide valuable insight to the mechanism
of Notch activation.
hN2 HD WT
hN2 HD L1573P
hN2 HD L1625P . MW IPTG IPTG
Ni Elution -IPTG IPTG Ni Elution
IPTG IPTG Ni Elution
His-WT
WT
hN2 Position hN2 Identity PS NG PL NP
1586 E 28.93 19.01 41.32 10.74
1587 L 0.00 26.05 30.25 43.70
1588 M 3.36 0.00 0.00 96.64
1589 V 0.00 0.00 0.00 100.00
1590 Y 3.36 2.52 0.84 93.28
1591 P 0.00 0.00 12.50 87.50
1592 Y 0.85 0.00 0.00 99.15
1593 Y 23.68 2.63 1.75 71.93
1594 G 4.63 9.26 2.78 83.33
1595 E 11.22 33.67 44.90 10.20
1596 K 61.26 23.42 10.81 4.50
1597 S 5.45 25.45 43.64 25.45
1598 A 1.02 21.43 26.53 51.02
1599 A 0.91 30.91 21.82 46.36
1600 M 7.27 1.82 13.64 77.27
1601 K 50.00 25.00 16.07 8.93
1602 K 61.61 23.21 10.71 4.46
1603 Q 24.32 1.80 57.66 16.22
1604 R 37.84 13.51 9.01 39.64
1605 M 8.04 1.79 4.46 85.71
1606 T 31.19 2.75 28.44 37.61
1607 R 67.54 3.51 15.79 13.16
1608 R 45.95 8.11 27.93 18.02
1609 S 10.91 4.55 50.91 33.64
1610 L 11.01 0.00 5.50 83.49
1611 P 3.81 2.86 5.71 87.62
1612 G 26.53 5.10 7.14 61.22
1613 E 12.75 50.98 21.57 14.71
1614 Q 10.00 18.00 60.00 12.00
1615 E 18.00 52.00 17.00 13.00
1616 Q 15.69 5.88 54.90 23.53
1617 E 9.43 61.32 8.49 20.75
1618 V 1.87 0.00 18.69 79.44
1619 A 25.00 0.89 23.21 50.89
1620 G 0.86 0.00 0.00 99.14
1621 S 0.00 0.00 64.96 35.04
1622 K 20.51 0.00 19.66 59.83
1623 V 0.00 0.00 1.71 98.29
1624 F 4.27 0.00 0.85 94.87
1625 L 0.00 0.00 0.85 99.15
1626 E 0.00 81.20 1.71 17.09
1627 I 0.00 0.00 0.00 100.00
1628 D 0.00 99.15 0.85 0.00
1629 N 0.00 0.00 85.59 14.41
1630 R 81.36 0.00 17.80 0.85
1631 Q 45.76 0.00 41.53 12.71
hN2 Position hN2 Identity PS NG PL NP
1632 C 0.00 0.00 100.00 0.00
1633 V 2.91 0.97 27.18 68.93
1634 Q 10.58 24.04 56.73 8.65
1635 D 1.02 36.73 44.90 17.35
1636 S 4.17 0.00 76.04 19.79
1637 D 0.00 42.71 46.88 10.42
1638 H 30.21 21.88 45.83 2.08
1639 C 0.00 0.00 89.32 10.68
1640 F 0.00 0.00 0.00 100.00
1641 K 27.18 17.48 38.83 16.50
1642 N 4.85 12.62 70.87 11.65
1643 T 0.00 0.00 30.69 69.31
1644 D 4.95 41.58 45.54 7.92
1645 A 0.99 44.55 16.83 37.62
1646 A 0.00 0.00 0.00 100.00
1647 A 0.00 0.00 0.00 100.00
1648 A 3.96 24.75 5.94 65.35
1649 L 0.99 0.00 10.89 88.12
1650 L 0.00 0.00 0.00 100.00
1651 A 0.99 0.00 0.99 98.02
1652 S 10.89 0.00 19.80 69.31
1653 H 32.67 1.98 13.86 51.49
1654 A 1.98 1.98 24.75 71.29
1655 I 0.00 0.00 41.00 59.00
1656 Q 30.30 0.00 34.34 35.35
1657 G 11.24 16.85 7.87 64.04
1658 T 11.63 3.49 72.09 12.79
1659 L 0.00 0.00 2.33 97.67
1660 S 5.62 12.36 68.54 13.48
1661 Y 0.00 0.00 0.00 100.00
1662 P 1.12 0.00 0.00 98.88
1663 L 0.00 0.00 0.00 100.00
1664 V 19.54 26.44 2.30 51.72
1665 S 2.30 16.09 49.43 32.18
1666 V 0.00 0.00 1.15 98.85
1667 V 28.00 1.33 32.00 38.67
1668 S 1.33 1.33 82.67 14.67
1669 E 4.17 88.89 6.94 0.00
1670 S 2.99 1.49 58.21 37.31
1671 L 4.92 6.56 0.00 88.52
1672 T 1.69 54.24 32.20 11.86
1673 P 0.00 0.00 0.00 100.00
1674 E 38.89 22.22 13.89 25.00
1675 R 52.94 0.00 26.47 20.59
1676 T 0.00 0.00 93.33 6.67
1677 Q 8.00 8.00 84.00 0.00
hN2 Position hN2 Identity PS NG PL NP
1540 E 0.00 98.25 1.75 0.00
1541 N 51.61 3.23 45.16 0.00
1542 L 0.00 0.00 0.00 100.00
1543 A 0.00 0.00 0.00 100.00
1544 E 12.00 57.00 3.00 28.00
1545 G 2.00 2.00 0.00 96.00
1546 T 1.68 0.84 63.03 34.45
1547 L 0.00 0.00 0.00 100.00
1548 V 0.84 0.84 19.33 78.99
1549 I 0.00 0.00 0.00 100.00
1550 V 0.84 0.84 5.04 93.28
1551 V 0.00 0.00 0.84 99.16
1552 L 10.92 0.00 1.68 87.39
1553 M 0.00 0.00 1.68 98.32
1554 P 11.76 15.13 21.85 51.26
1555 P 0.00 0.00 0.83 99.17
1556 E 8.33 63.33 8.33 20.00
1557 Q 2.50 19.17 52.50 25.83
1558 L 0.00 0.00 0.00 100.00
1559 L 45.83 16.67 1.67 35.83
1560 Q 15.00 8.33 67.50 9.17
1561 D 4.17 25.00 58.33 12.50
1562 A 2.50 0.00 56.67 40.83
1563 R 23.33 0.00 8.33 68.33
1564 S 18.18 5.79 42.98 33.06
1565 F 0.00 0.00 0.00 100.00
1566 L 0.00 0.00 0.00 100.00
1567 R 89.26 0.00 9.92 0.83
1568 A 8.26 40.50 14.88 36.36
1569 L 0.00 0.00 0.83 99.17
1570 G 0.00 0.83 68.60 30.58
1571 T 44.63 0.00 26.45 28.93
1572 L 1.65 3.31 17.36 77.69
1573 L 0.00 0.00 0.00 100.00
1574 H 95.87 0.83 2.48 0.83
1575 T 0.00 0.00 82.64 17.36
1576 N 0.83 0.00 79.34 19.83
1577 L 0.83 0.00 0.00 99.17
1578 R 46.28 0.83 5.79 47.11
1579 I 0.83 0.00 0.00 99.17
1580 K 98.35 0.00 0.00 1.65
1581 R 71.67 0.83 9.17 18.33
1582 D 1.67 90.00 8.33 0.00
1583 S 4.20 9.24 43.70 42.86
1584 Q 23.33 21.67 36.67 18.33
1585 G 1.65 0.83 4.13 93.39
22 kDa

• q

(B.)
(A.)
Figure 2. hN2 WT, L1573P, and L1625P
purification. (A.) Expression and nickel affinity
purification of recombinant hN2 HD domains
(wildtype WT and two mutants L1573P, L1625P) in
E. Coli. IPTG cell lysate prior to IPTG
induction IPTG cell lysate four hours
post-IPTG induction Ni Elution Protein sample
eluted from a nickel affinity chromatography
column MW Protein molecular weight marker.
Differences in band intensity reflect variations
in volumes of sample loaded. migration position
of the expressed HD domain. (B.) hN2 HD WT C18
HPLC purification. The identity of the different
peaks was confirmed by mass spectrometry
analysis.
(A.)
90 rotation
Conclusions and Future Directions
Modeling Our modeling results showed no
significant structural differences between the
hN2 HD WT and the mutants. However, our modeled
HD WT domain had a significant backbone RMSD from
the template (RMSD 0.49). Therefore, we will
repeat modeling with additional energy
minimization followed by molecular dynamics
protocols. CD data Our preliminary results
indicate qualitative secondary structure
differences between WT and the mutants, L1573P
and L1625P. We will complete quantitative
analysis of this CD scan data to evaluate the
significance of these differences. Since the
multiplicity of steps in the protein purification
protocol resulted in very low protein yields and
very dilute protein concentrations for performing
CD experiments, we are currently expressing and
purifying the His-tagged versions of these
proteins to repeat the experiments. In addition,
we only had enough protein to perform the thermal
melt for L1573P, which showed a significant
degree of destabilization (?Tm 18 C) when
compared to HD WTdel. We will also repeat the
unfolding experiments using the His-tagged
versions of WT, L1573P, and L1625P. Amino acid
conservation and stability Our table indicates
that many positions along the HD sequence are
highly conserved. When these positions are mapped
onto the HD structure (Figure 6), they correspond
very well with the residues that make up the
hydrophobic core (Figure 3C). Based on L1573P and
L1625P experimental findings and calculation of
prevalence of physicochemical properties we will
select further mutations to test experimentally.
(B.)
(C.)
Figure 3 (A.) Swissmodel and GROMACS modeling of
the wildtype HD domain of Notch 2 (WT). Mutated
amino acids L1573 and L1625 are shown in dark
blue stick model. (B.) Superposition of L1573P
and L1625P, the two mutants used in this
experiment with WT. Residues L1573 and L1625 in
WT are in dark blue line model in mutants
proline point mutations are in magenta stick
model. Modeling results did not reveal
significant perturbations to the secondary
structure around the mutations or elsewhere in
the HD domain. (C.) Same as (B.) with most of the
hydrophobic core of the HD domain highlighted in
cyan. Amino acids found to have less than 10
solvent accessibility, including L1573 and L1625
are in dark blue and L1573P and L1625P in
magenta, shown in space-fill model. A few of the
hydrophobic core residues are not displayed for
clarity of the location of the two leucines under
investigation. L1573 and L1625 partake in the
hydrophobic core of the HD domain, and are thus
predicted to be crucial to the maintenance of its
stability.
Figure 1. (A.) Overview of Notch signaling with
the domain organization of the Negative
Regulatory Region (NRR) bracketed in pink.3 (B.)
X-Ray diffraction structure of the Negative
Regulatory Region (NRR) of Notch with the
Heterodimerization (HD) domain bracketed orange.3
(C.) Energy minimized model of the wildtype HD
domain of Notch2 (see Molecular Modeling and
Energy Minimization under Computational
Methods).4,5,6
Methods
Experimental
Computational
Protein expression Recombinant hN2 HD wildtype
(WT), and L1573P and L1625P domains were
expressed as N-terminally His-tagged protein
using a modified pET21a vector that contained a
TEV cleavage site between the N-terminal His-tag
and the protein of interest. Protein purification
Expressed protein was solubilized and separated
from insoluble components through several
sonication and centrifugation steps. Initial
purification was performed through nickel
affinity chromatography. His-tag was then removed
via overnight TEV cleavage. Final purification
was performed by C18 HPLC. Purified protein
identity was confirmed by MS. Circular dichroism
(CD) Measurements Spectra are the average of
five scans taken at 20 C between 198 and 260 nm,
with 1 nm steps, using 1 nm bandwidth and
averaging time 5 s for each wavelength in a 0.1cm
cuvette. All spectra were baseline corrected
against the same buffer scan. Circular dichroism
(CD) Thermal Unfolding Thermal unfolding was
monitored by CD signal at 217 and 222 nm. Protein
samples were prepared identical to those for the
CD scans. Thermal unfolding was performed from
4 C to 90 C, in 1 C increments using an
equilibration time of 1 min at each temperature
step, followed by a data-averaging time of 25 s
at each degree in a 1 cm cuvette.
Molecular Modeling Default SWISSMODEL4 settings
were used for modeling WT, L1573P, and L1625P.
The X-Ray diffraction structure of hN2 NRR (PDB
ID 2OO4.pdb) was used as the template
structure. All structures are visualized using
PyMOL5. Energy Minimization Energy minimizations
were performed on WT, L1573P, and L1625P using
the conjugate gradient method. Coulombic and van
der Waals interactions were cut off at 1.0 nm.
Analyses were performed using the tools available
in the GROMACS suite.6 BLAST Default BLAST
settings were used to identify 126 sequences that
are common to hN2, hN1, and hN3 HD BLAST
results.7 Multiple Sequence Alignment The 126
homologous sequences were aligned based on BLAST
results to hN2 HD. In order to maintain the
alignment with the 138 amino acids of hN2 HD, the
insertions encountered upon comparison of the 126
homologous proteins and hN2 were excised.7 Python
A Python program was developed for determining
the prevalence of positive (PS), negative (NG),
polar (PL), and nonpolar (NP) amino acids at each
amino acid position. For this work, PS R, H,
and K NG D and E PL S, T, N, Q, and C and
NP G, P, A, V, I, L, M, F, W, and Y.8
Figure 6. For every amino acid position in hN2 HD
WT, mapping of prevalence of physicochemical
properties to hN2 HD structure. See Table 1 for
more details. 4,5,6
Acknowledgements
We would like to thank the Blacklow Laboratory at
Brigham and Womens Hospital/Harvard Medical
School for Circular Dichroism Instrument time and
the Wellesley College Chemistry Department for
supporting our research through the Beck
Fellowship.
References
Tm 80
Tm 62

• Malecki, M.J., et al., Leukemia-associated
  mutations within the NOTCH1 heterodimerization
  domain fall into at least two distinct
  mechanistic classes. Molecular and Cellular
  Biology, 2006. 26(12) p. 4642-4651.
• Weng, A.P., et al., Activating mutations of
  NOTCH1 in human T cell acute lymphoblastic
  leukemia. Science, 2004. 306(5694) p. 269-271.
• Gordon, W.R., et al., Structural basis for
  autoinhibition of Notch. Nature Structural
  Molecular Biology, 2007. 14(5) p. 455-455.
• Arnold K., et al., The SWISS-MODEL Workspace A
  web-based environment for protein structure
  homology modelling. Bioinformatics, 2006. 22
  pp.195-201. http//swissmodel.expasy.org//SWISS-MO
  DEL.html
• DeLano W.L., The PyMOL Molecular Graphics System,
  2002. http//www.pymol.org
• Hess B., et al., GROMACS 4 Algorithms for highly
  efficient, load-balanced, and scalable molecular
  simulation, J. Chem. Theor. Comp., 2008. 4 pp.
  0. http//www.gromacs.org/
• Gertz, E.M., BLAST Scoring Parameters, 2005.
  http//blast.ncbi.nlm.nih.gov/Blast.cgi
• Python Software Foundation, Python 2.5.1, 2007.
  http//www.python.org/

Figure 5. Comparison of the thermal unfolding of
L1573P with HD Wtdel monitered at 217 and 222
nms. HD WTdel is a WT hN2 HD domain with an
excised furin loop. Thermal unfolding midpoints
(Tm) were estimated as the midpoint between the
unfolded and folded baselines shown in the figure
with dashed lines. The Tm determined from the
ellipticity at two wavelengths (217 and 222 nms)
for each construct is the same. Compared to HD
WTdel, L1573P mutant is destabilized (?Tm 18C)
consistent with our modeling results showing
tight packing of these residues within the
hydrophobic core.
Figure 4. Superposition of the circular dichroism
spectra from the hN2 HD WT, and L1573P and L1625P
mutants. Protein samples (10-40 uM) were
resuspended in 25 mM phosphate buffer at pH 7.0
and 100 mM NaCl. Spectra were corrected for
concentration differences across samples.
Preliminary results indicate qualitative
differences in secondary structure among the
three HD domains studied. Quantitative analysis
is ongoing.