Transcript Slide 1

Characterization of protein folding determinants for LIN-12/Notch-Repeats (LNRs)
using Human Notch1 LNR-B as a model system
Sharline Madera
Advisor: Dr. Didem Vardar-Ulu
Wellesley College, Massachusetts
NSF REU Chemistry
Leadership Group
Travel Award
Recipient
NSF Award CHE-0754512
LNR Domain
A B C
HD Domain
Metal Specificity
5 Constructs
Disulfide Connectivity
Human Notch1 is a member of a conserved family of heterodimeric type 1 transmembrane receptors that
control differentiation in multicellular animals. Notch proteins possess three contiguous LIN-12/NotchRepeats (LNRs), LNRA, LNRB and LNRC, in their extracellular domain that maintain the receptor in its
resting conformation in the absence of ligand. These conserved LNRs display a characteristic disulfide
bonding pattern and require Ca2+ for folding. In the receptor, they are separated by two linkers, linker_AB
and linker_BC, 10 and 6 amino acids long, respectively. Previously, we had shown that LNRB folding was
affected by residues 6-10 of linker_AB and not of linker_BC (1, Figure 5). In this study we investigated the
dependence of LNRB folding on the identity of the metal ion as well as the number of conserved disulfide
bonds. Our results indicate that LNRB folding is selective for Ca2+ and the linker_AB length requirement
and total number of disulfide sulfide bonds needed for effective folding are interdependent. Together
these studies represents the initial steps toward defining the minimum requirements for a correctly folding
LNR module using LNRB from human Notch1 as a model system.
C
C
SLNFNDPWKN
C
C
QRAEGQ
C
2:4
3:6
LNRB_orig:
Mut 1,5C→A LNRB_orig:
LNRB_short:
LNRB_delAB:
LNRB_delBC:
Mut 1,5C→A LNRB_delBC:
LNRB_int:
Mut 1,5C→A LNRB_int:
MgCl2
LNFNDPW KNATQ SLQCWKYFSDGH CDSQ CNSAG ALFDGFDC QRAEG Q
KN CTQ SLQ CWKYFSDG HCDSQCNSAG CLFDG FDCQ
KN CTQ SLQ CWKYFSDG HCDSQCNSAG CLFDG FDCQ RAEG Q
LNFNDPWKNCTQSLQCWKYFSDGHCDSQCNSAGCLFDGFDCQ
LNFNDPWKNATQSLQCWKYFSDGHCDSQCNSAGALFDGFDCQ
DPWKNCTQSLQCWKYFSDGHCDSQCNSAGCLFDGFDCQ
DPWKNATQSLQCWKYFSDGHCDSQCNSAGALFDGFDCQ
Figure 7. Chromatograms of LNRB_orig and LNRB_int folded under varying metal ion conditions.
Panel A: Superposition of chromatograms of LNRB_orig folded in the presence of 10mM CaCl2-red and
10mM MgCl2-black. Panel B: Superposition of chromatograms of LNRB_orig folded in the presence of 1mM
CaCl2-green and 1mM ZnCl2-blue. These chromatograms demonstrate the selectivity of LNRB_orig folding for
Ca2+ which result in a single folded peak, compared to other potential divalent cations Mg2+ and Zn2+, which
yield an array of multiple peaks indicating the lack of one predominant native fold.
LNRA
LNRC
LNRB
Figure 1. Human Notch 1 LNRs and linkers.
Table 1. Summary of HPLC & Mass Spectrometry Results
LNRB_orig
Post DTT
Post DTT
LNRB_int
LNRB_delBC
LNRB_delAB
LNRB_short
B
A
Figure 5. Chromatograms of LNRB constructs: LNRB_orig- green, LNRB_int- purple, LNRB_delBC- blue. LNRB_delABbrown, LNRB_short- orange Panel A) Superposition of the chromatograms of the three LNRB constructs that folded into a single
predominant peak indicating a single thermodynamically favored correctly folded species. Small neighboring peaks are indicative
of misfolded species. Top left inset: Representative chromatogram detailing the elution gradient used and the pressure during the
run. Panel B) Superposition of the chromatograms of the two unfolded constructs after dialysis 3. Note no predominant peak is
obtained demonstrating no preference for correctly folded species for these two constructs. Top right insets for both panels:
Superposition of chromatograms for the same constructs after DTT treatment showing that all peaks of the folded chromatograms
collapse to a single peak when reduced and elute later in the gradient (see Table 1).
LNRB_delBC
Mutant
LNRB_delBC
Figure 2. Domain organization of the Notch Receptor.
Figure 3. Crystal structure of Human Notch2 NRR (15).
Figure 6. Superposition of chromatograms for LNRB_delBC-black and Mutant LNRB_delBC-red. Similar to LNRB_delBC,
mutant LNRB_delBC displays a single predominant peak. This finding demonstrates that the mutant LNRB_delBC that contains
only two disulfide bonds is capable of folding into a single thermodynamically favored state whose stability is comparable to that of
wild type LNRB_delBC.
• Wildtype hN1LNRB was expressed as inclusion bodies in • Soluble LNRB constructs (~175 M) were folded for two days
BL21(DE3)PlysS E.coli. cell line as a fusion protein with a in a refolding buffer with daily buffer changes.
modified form of the trpLE sequence using the pMML vector
100mM NaCl
(kind gift of S. Blacklow, BWH).
 20mM Tris pH 8
10mM CaCl2/ 10mM MgCl2/ 1mM ZnCl2
• The plasmids for the mutant hN1LNRBs were obtained
2.5mM cysteine
from the corresponding wildtype pMML vector using the
 0.5mM cystine
QuikChange
Site-Directed
Mutagenesis
protocol
(Stratagene) where cysteines 1 and 5 were changed into • On day 3 the constructs were moved into a dialysis buffer that
alanines eliminating the potential for the first disulfide bond. did not contain any redox reagent (cysteine/cystine).
•LNRB was cleaved from the hydrophobic leader sequence • All constructs from day 3 of dialysis were analyzed on a
by cyanogen bromide cleavage in 70% formic acid and was reverse phase HPLC using a C18 column and 0.25%/min
separated from the leader sequence through precipitation of gradient elution under the following experimental conditions:
Buffer A: 10% Acetonitrile, 90% H2O, 0.1% TFA
the leader sequence upon pH increase.
Buffer B: 90% Acetonitrile, 10% H2O, 0.1% TFA
• Protein identity for each construct was confirmed by
• A sample of each folded construct was also incubated with
Mass Spectrometry.
100 mM DTT at room temp for 2 hrs and analyzed by RP-HPLC.
CaCl2
LNFNDPW KNCTQ SLQCWKYFSDGH CDSQ CNSAG CLFDGFDCQRAEG Q
C
Notch Proteins are large Ca2+ binding, transmembrane receptors that control differentiation in multicellular
animals. In mammals, there are four Notch homologs: Notch1-4. These proteins function via a highly
conserved mechanism referred to as the Notch signaling pathway, which is important for cell-cell
communication, involving gene regulation mechanisms that control multiple cell differentiation processes
during embryonic and adult life. Deregulation of normal Notch activation has been noted in certain human
leukemias, (2) Alagille (3, 4) and CADASIL (5) syndromes, indicating that perturbations of Notch signaling
underlie several forms of human diseases (6). Notch proteins exhibit a highly conserved modular
architecture (Figure 2), in which distinct repeated structural units are associated with different functional
roles in the intact receptor (7). Ligand binding to the N-terminal EGF-repeats activates these proteins by
facilitating a proteolytic cleavage by a metalloprotease at site S2, and triggers the gamma-secretase
cleavage that permits the translocation of intracellular Notch (ICN) into the nucleus to activate the
transcription of target genes (8, 9, 10, 11). The Negative Regulatory Region (NRR) of all Notch receptors
has three tandem, independently folding LIN-12/Notch Repeats (LNRs) that wrap around the HD domain
containing the regulatory cleavage site S2, and mask the S2 site in the resting receptor (Figure 3) (12-14).
Hence the interactions between the LNRs and the HD are critical in stabilizing the NRR and preventing
activation prior to ligand binding. Each of the LNRs contains six cysteines with a unique disulfide bonding
pattern and coordinate a single Ca2+ (Figure 3), however the minimum requirements that would ensure an
LNR to fold independently are not known. This work utilizes different LNRB constructs that all contain the
32 amino acid stretch from cysteine 1 to cysteine 6 in the second LNR of hN1and various numbers of the
residues that flank these residues (Table 1), to define the minimum length requirement for hN1 LNRB and
to investigate the impact of metal ions and number of disulfide bonds on its autonomous folding.
B
ZnCl2
Figure 4. Sequence Alignment for LNRB Constructs used for the study: Red Residues: Ca 2+ coordinating residues
Blue Residues: Mutated cysteines to alanines Orange Residues: Disulfide bonded cysteines
linker_BC
linker_AB
1:5
CaCl2
A
A
B
LNRB_orig
Mutant
LNRB_orig
Construct
% Buffer B
Elution
% Buffer B Elution
100mM DTT
Correctly Folded
Calculated MW
(Da)
Mass Spec MW
(Da)
LNRB_orig
28
30
Yes
5368.8
5365.65
Mut LNRB_orig
28.2-30.2
---
No
5304.7
---
LNRB_short
21-25
26
No
3940.3
3937.54
LNRB_int
25
27
Yes
4338.7
4291.93
Mut LNRB_int
22.7-32.3
---
No
4274.6
4227.36
LNRB_delAB
23-25
25
No
4481.8
4480.47
LNRB_delBC
27
30
Yes
4827.2
4823.22
Mut LNRB_delBC
26.4
---
Yes
4763.1
4760.86
In this study we investigated the relative importance of the number of potential disulfide bonds and the
divalent ion identity for the independent folding of LNRB in relation to the presence or absence of the linker
regions flanking LNRB. Previous folding studies identified constructs LNRB_delBC, LNRB_orig and LNRB_int
as autonomously folding constructs. In this study we show two of the modified LNRB constructs where
cysteines 1 and 5 are mutated to alanines to eliminate the potential of the first disulfide bond, Mut 1,5C→A
LNRB_int and Mut 1,5C→A LNRB_orig, are unable to fold into a single thermodynamically favored folded
state. In contrast, Mut 1,5C→A LNRB_delBC retains the ability to fold autonomously. These data show that
for the formation of a single thermodynamically favored LNRB species upon in vitro refolding there is a
minimum requirement for the total number of stabilizing interactions that participate in the folding process.
The relative impact of each of these interactions, which involve linker_AB, the disulfide bonds, and the Ca2+
ion, in ensuring autonomous folding are highly interdependent. Furthermore the attainment of one
thermodynamically favored species in the folding of LNRB is dependent and selective for Ca2+ compared to
the two other divalent cations abundant in cells, Mg2+ and Zn2+.
The findings of this study are the initial steps in defining the minimum requirements for an autonomously
folding LNR module using LNR_B as a model from human Notch1. They will be followed by studies that
include further investigations on the effects of altering the cysteine arrangement of other LNR modules with
varying lengths and direct correlation of these findings to hN4LNRA, a wild type LNR module within the human
Notch4 receptor and contains only two disulfide bonds.
LNRB_int
Mutant
LNRB_int
Figure 7. Panel A) Superposition of the chromatograms of LNRB_orig-black and Mutant LNRB_orig-red. Panel B)
Superposition of the chromatograms of LNRB_int-black and Mutant LNRB_int-red. Unlike the wild type constructs LNRB_orig
and LNRB_int, mutant LNRB_orig and mutant LNRB_int do not display a single peak that corresponds to one thermodynamically
stable folded state. Instead the chromatograms of the mutant constructs show various peaks indicating the presence of several
thermodynamically similar species and not one favored folded state.
1. Madera, S. and Vardar-Ulu D. “Characterization of protein folding determinants for LIN-12/Notch-Repeats
(LNRs) using Human Notch1 LNR-B as a model system” Poster Presentation, 21st Annual Symposium of
the American Protein Society, Protein Society, Boston July 2007.
2. Ellisen, L. W. et al. (1991) Cell. 66:649–661.
3. Li, L., et al. (1997) Nat. Genet. 16: 243–251.
4. Oda, T. et al. (1997) Nat. Genet. 16: 235–242.
5. Joutel, A. et al. (1996) Nature. 383: 707–710.
6. Rand, M. et al. (2000) Molec. and Cell. Biol. 20: 1825-1835.
7. Vardar, D. et al. (2003) Biochemistry. 42: 7061-7067.
8. Sanchez-Irizarry, C. et al. (2004) Molec. and Cell. Biol. 24: 9265-9273.
9. Logeat, F. et al. (1998) Proc. Natl. Acad. Sci. USA. 95: 8108-8112.
10. Brou, C. et al. (2000) Mol. Cell. 2: 207-216.
11. Lawrence, N. (2000) Development. 127: 3185-3195.
12. Aster, J. et al. (1999) Biochemistry. 38: 4736-4742.
13. Weng, A.P. et al. (2004) Science. 306: 269–271.
14. Kopan, R. et al. (2000) Genes Dev. 14: 2799-2806.
15. Gordon, W. R. et al. (2007) Nature Structural and Molec. Biol. 14: 295-300.