ISOLATION AND PURIFICATION OF HIGH MOBILITY GROUP-1 PROTEIN AND ITS EFFECT ON THE BINDING OF ESTROGEN RECEPTORS TO THE ESTROGEN RESPONSE ELEMENT

1

• ISOLATION AND PURIFICATION OF HIGH MOBILITY
  GROUP-1 PROTEIN AND ITS EFFECT ON THE BINDING OF
  ESTROGEN RECEPTORS TO THE ESTROGEN RESPONSE
  ELEMENT
• Richard Hausman
• Summer 2006

2
ABSTRACT

• The high mobility group box (HMGB) proteins are
  abundant, ubiquitous and highly conserved
  proteins that play a major role in transcription.
  HMGB proteins, especially HMGB-1, are of
  interest in this study due to their effect on the
  binding of nuclear hormone receptors to their
  response elements. For this reason, HMGB-1 was
  isolated and purified. Once the purified protein
  was obtained, electrophoretic mobility shift
  assays (EMSA) were performed using different
  nuclear hormone receptors with their cognate
  response elements in the presence and absence of
  HMBG-1. The nuclear hormone receptors of
  interest include estrogen receptor ? and estrogen
  receptor ?, both of which were used with their
  consensus response element (cERE) during EMSAs.
  The consensus sequence to which ER binds is
  AGGTCAnnnTGACCT. These studies showed that
  HMGB-1 had an enhancing effect on the binding of
  ER?/ER? to cERE, reducing the KD values for both
  estrogen receptors by a factor of 3.5.

3
I. INTRODUCTION

• A. Nuclear receptors Nuclear receptors are a
  super family of ligand-activated transcription
  factors that modulate specific gene expression.
  They are categorized into three steroid receptor
  subfamilies. The focus of this research has been
  on Class I steroid receptors, which includes
  estrogen receptors (ER) glucocorticoid receptors
  (GR), mineralocorticoid receptor (MR), androgen
  receptor (AR), and progesterone receptor (PR) and
  are classically defined as ligand-dependent and
  act as homodimers (Weatherman, Flletterick, and
  Scanlan, 1999).

4

• Nuclear receptors (cont) Steroid receptors
  reside in either the cytoplasm or the in the
  nucleus and are bound to a number of chaperones
  such as heat shock proteins and immunophillins.
  Upon binding and consequently, activation, by
  their hormones in the cytoplasm, the steroid
  receptors dissociate from their chaperones and
  dimerize. This dimerization occurs through
  protein-protein interactions between the
  receptors dimerization domains. The dimmer then
  targets specific sequences on DNA, call hormone
  response elements (HRE), in the nucleus and
  either activates or represses the transcriptional
  machinery. Transcription is mediated by
  interaction of co-activators, co-repressors and
  components of the pre-initiation complex.

5

• B. Estrogen Receptors Estrogen has
  multifunctional effects on growth,
  differentiation, and function in many tissues.
  It can diffuse in and out of cells, but is
  retained in the target cell nuclei by estrogen
  receptor (ER). Once estrogen is bound, the ER
  undergoes conformational changes allowing the
  receptor to interact with its sequence specific
  estrogen response element (ERE) on DNA. This
  allows it to modulate the expression of target
  genes.

6

• C. High Mobility Group Box-1/-2 (HMGB-1/-2)
  Proteins The high mobility group box (HMGB)
  non-histone chromosomal proteins were first
  discovered and characterized by Goodwin and Johns
  in 1982. They were found in mammalian cells and
  named according to their movement in acid urea
  gels, since their function was still unknown.
  HMGB proteins are chromosomal proteins that are
  extractable from chromatin in 0.35 M NaCl, are
  soluble in 2-5 perchloric or trichloroacetic
  acid, have a high content of charged amino acids
  (20-30), are relatively high in pralines, and
  have a relatively high electrophoretic mobility
  in polyacrylamine gels with a molecular weight
  lower than 30 kDa (25kDa) (Johns, 1982). HMGB
  proteins can bind to both double stranded and
  single stranded DNA, but prefer the former
  (Johns, 1982 Eink et al.,1985). Several
  experimental evidences suggest that HMGB proteins
  play a role in chromosomal replication (Bonne et
  al., 1982) Duguet et al., 1978).

7

• High Mobility Group Box-1/-2 (HMGB-1/-2)
  Proteins (cont) The negatively charged
  C-terminal region of HMGB-1 is thought to be the
  protein region which facilitates DNA unwinding
  (Yoshida et al., 1984). Antibody inhibition
  studies have suggested that HMGB-1 may have some
  function in DNA replication (Alexandrova et al.,
  1987). HMGB-1 may regulate gene transcription by
  helping other factors to bind DNA and activate
  RNA polymerase. Although HMGB-1 has been known
  to act as a negative cofactor in transcriptional
  regulation (Ge et al., 1994), it has been shown
  to interact in vitro to enhance the
  transcriptional activities of several
  transcription factors including GR, ER, PR,
  homeobox proteins, POU domain transactivators,
  TBP, p53, etc. (Zwilling et al., 1995 Zappavigna
  et al., 1996 Prendergast et al., 1994
  Boonnyakaratanakornkit et al., 1998 Melvin et
  al., 1999 Onate et al., 1994). Mice that were
  HMGB1 deficient show a distinct phenotype and
  died shortly after birth due to hypoglycemia and
  were deficient in the activation of GR responsive
  genes (Calogero et al., 1999). This supports the
  idea that HMGB-1 contributes to gene specific
  transcriptional regulation. Furthering the
  understanding of HMGB-1s role in transcription,
  it was found that HMGB-1 couldnt bind alone to
  ERE but still enhanced the binding of ERa to
  cERE. This suggested a protein-protein
  interaction between ER and HMGB-1 (Das et al,
  2004).

8

• Research Objectives
• My study focused on these aspects
• 1. Isolate and Purify HMGB-1 from calf thymus
• 2. Binding studies involving the binding of
  ERa/ERß to 33bp cERE
• 3. Show the effect of HMGB-1 on the binding
  affinities of ERa/ERß to 33bp cERE

9
HMGB-1/HMGB-2 Isolation and Purification Results

• HMGB-1/HMGB-2 Isolation and Purification
• The HMGB-1 and HMGB-2 proteins were
  isolated from calf thymus under denaturing
  conditions consisting of salt extraction and
  ammonium salt precipitations. This crude
  isolation resulted in a solution that was
  approximately 50 HMGB-1/HMGB-2 as Figure 1
  shows. Purification was performed using an HPLC
  unit and a buffered linear salt gradient from 0.2
  M NaCl to 0.8 M NaCl on a Mono Q 5/5 anion
  exchange column. A 400µL volume of the
  HMGB-1/HMGB-2 concentrated solution (HMGB conc.)
  was run on the column using the HPLC unit in
  order to obtain a chromatogram. The chromatogram
  has 4 distinct peaks and 3 smaller areas of
  absorption (Figure 2). The samples collected
  from these seven areas of absorption were run on
  a 12 polyacrylamide SDS-PAGE with an aliquot of
  the HMGB conc. in order to determine which peaks
  belonged to HMGB-1 and HMGB-2 (Figure 3). Once
  it was determined which band corresponded with
  HMGB-1 and HMGB-2, fractions were collected at
  the corresponding retention times. To ensure
  purity, it was made sure that the peaks were not
  overlapping (Figure 4).
• HMGB-1 from calf thymus contains 215
  amino acids and has a molecular weight of 25 kDa
  (Chow et al., 1995). After purification by HPLC,
  the solutions were concentrated and their purity
  was determined using SDS-PAGE. Both proteins
  migrate as a single protein with a molecular
  weight of 25 kDa, but on a 12 polyacrylamide gel
  in SDS-PAGE two distinguishable bands are present
  due to HMGB-2 migrating faster than
  HMGB-1(Goodwin, et al., 1979). Although the
  amounts of HMGB-2 were less than those of HMGB-1,
  the purity of both proteins was well over 90
  (Figure 5).
• To determine the concentration of the
  HMGB-1 solution it was run in a spectrophotometer
  and the UV absorption at 280 nm was obtained.
  Using this absorption (Figure 6), the Beer-
  Lambert Law, and an extinction coefficient of
  20,500/Mcm (Chow et al.,1995) it was found that
  the concentration of 1073 ng/µL . The yield that
  was obtained was 1.05 mg of HMGB-1 for 150 g of
  calf thymus. Since the HMGB-2 was not going to
  be used in EMSA studies, it was purified with the
  HPLC, but was not analyzed after that point.

10

• Figure 1 SDS PAGE Analysis of HMGB Solution
  After Crude Isolation The solution was analyzed
  on a 12 polyacrylamide gel. Lanes 1 and 4
  contain the crude isolated HMGB-1/HMGB-2 solution
  at 5µL and 10µL, respectively. Lanes 2 and 5
  contain a molecular weight standard at 5µL and
  10µL, respectively. Lanes 3 and 6 contain 2.5µL
  of pure HMGB-1 at 951 ng/µL. The majority of
  the protein in lanes 1 and 4 are that of
  HMGB-1/-2.

11

• Figure 2 HPLC Chromatogram of 400µL Crude HMGB
  Isolated Solution The red lines indicate where
  the collections were started and the black lines
  indicate where the collections were ended.
  Collections 3 and 4 represent HMGB-2 and HMGB-1
  respectively.

12

• Figure 3 SDS-PAGE Analysis of the Chromatogram
  Peaks Obtained from HPLC Purification of Isolated
  HMGB Solution Lanes 1 through 7 contained 5µL of
  their respective peaks while the eighth lane
  contained 10µL of the crudely isolated HMGB
  solution. Since the isolated HMGB solution had
  been run against HMGB-1 on a previous SDS-PAGE
  gel, the location of the HMGB-1 and HMGB-2 band
  were known. Peak 3 and Peak 4 of the chromatogram
  contained HMGB-1 and HMGB-2 respectively.

13

• Figure 4 HPLC Chromatogram Indicating
  HMGB-1/HMGB-2 Peaks HMGB-2 eluted before HMGB-1
  with a retention time of 10.8 minutes compared to
  HMGB-1s retention time of 11.5 minutes. The red
  slash marks indicate where the samples were
  collected in order to maximize volume while
  retaining purity.

14

• Figure 5 12 Polyacrylamide SDS-PAGE Indicating
  the Purity of HMGB-1 Lane 1 contains 10 µL of
  the freshly purified HMGB-1, Lane 2 contains 5 µL
  of the freshly purified HMGB-1, Lane 3 contains
  10 µL of the old purified HMGB-1, Lane 4 contains
  10 µL of the freshly purified HMGB-2, Lane 5
  contains 5 µL of the freshly purified HMGB-2, and
  Lane 6 contains 10 µL of the molecular weigh
  marker. This shows that the purity of HMGB-1 and
  HMGB-2 is estimated at over 90.

15

• Figure 6 Absorbance Graph from
  Spectrophotometer Indicating the Absorbance of
  HMGB-1 at ? 280nm The arrow on the graph
  indicates the absorbance at a wavelength 280nm
  which is 0.880. Using the Beer-Lambert law, the
  concentration of the HMGB-1 solution was
  calculated to be 1073 ng/mL.

16
Comparison of ER? and ER?s binding affinity to
cERE with and without HMGB-1

• Table III KD value for ERa/ERß to cERE from
  Previous Studies

• The purpose of this study was to study the
  binding of ERa/ERß to 33bp cERE in the presence
  and absence of HMGB-1. It has been found in
  previous studies that the KD values of ERa/ERß to
  33bp cERE are as follows (Ghattamaneni, 2005)

17

• It was intended to reproduce results that would
  be similar to these that being results with
  similar KDs for ERa/ERß to 33bp cERE and a
  similar effect from HMGB-1. Figure 7 shows a
  titration of cERE with ERa in the presence and
  absence of HMGB-1. Increasing amounts of ERa
  were added to a constant amount of 32P labeled
  DNA probes and the complex was resolved on a 5
  polyacrylamide gel. Similarly, Figure 10 shows a
  titration of cERE with ERß in the presence and
  absence of HMGB-1. Increasing amounts of ERß
  were added to a constant amount of 32P labeled
  DNA probes and the complex was resolved on a 5
  polyacrylamide gel. The ERa/ERß forms a
  complex with the free DNA making a more slowly
  moving band. As the ERa/ERß concentration is
  increased, the amount of complex also increases
  and the amount of free DNA decreases. In the
  presence of HMGB-1 a complex between free DNA and
  ERa/ERß occurs at lower ERa/ERß concentrations
  and increases more rapidly with the same increase
  of ERa/ERß concentrations. This supports the
  idea that HMGB-1 enhanced the binding affinity of
  ERa/ERß and cERE. Figure 8 and Figure 9 show the
  equilibrium binding profiles of ERa to cERE in
  the presence and absence of HMGB-1, respectively.
  From these profiles, the KD values can be
  estimated at 7.4 nM and 2.1 nM in the presence
  and absence of HMGB-1 respectively. This shows
  an increase of ERa binding to cERE by a factor of
  3.5. Figure 11 shows the equilibrium binding
  profiles of ERß to cERE in the presence and
  absence of HMGB-1. Similar to ERa, the KD values
  from these profiles can be estimated at 7.5 nM
  and 2.1 nM in the presence and absence of HMGB-1
  respectively. This also shows an increase of ERß
  binding to cERE by a factor of 3.5.

18
w/o HMGB-1 w/ HMGB-1

• 1 2 3 4 5
  6 7 8 9 10 11
  12 13 14 15 16

Figure 7 Comparison of ERa binding to cERE in
the presence and absence of HMGB-1 A 100 pM
probe of cERE (lanes 1-16) was incubated with 0
nM (lanes 1 9), 0.5 nM (lanes 2 10), 1.0 nM
(lanes 3 11), 2.0 nM (lanes 4 12), 4.0 nM (
lanes 5 13) 8.0 nM (lanes 6 14) 16.0 nM
(lanes 7 15) and 32.0 nM (lanes 8 16) of ERa
and 400 nM HMGB-1 (lanes 9-16). In all lanes the
amount of complex increased while the amount of
free DNA decreased with increasing concentrations
of ERa. The increase in complex, and
consequently the decrease in free DNA, occurs at
lower concentration of ERa in the presence of
HMGB-1, supporting the idea that HMGB-1 enhances
the binding affinity of ERa to cERE.
19
Complex profile
ERa

• Figure 8 Equilibrium binding profile of ERa
  binding to cERE in the absence of HMGB1 The
  percent ERa complex bound in the absence of
  HMGB-1 obtained from EMSA profiles are plotted as
  a function of ERa. From this profile the KD for
  the binding of ERa to cERE can be estimated at
  7.4 nM.

20

• Figure 9 Equilibrium binding profile of ERa
  binding to cERE in the presence of HMGB1 The
  percent ERa complex bound in the presence of
  HMGB-1 obtained from EMSA profiles are plotted as
  a function of ERa. From this graph, the KD value
  of ERa binding to cERE can be estimated at 2.1 nM.

21
w/o HMGB-1 w/ HMGB-1

• 1 2 3 4 5
  6 7 8 9 10 11 12 13 14 15 16

Figure 10 Comparison of ERß binding to cERE in
the presence and absence of HMGB-1 A 100 pM
probe of cERE (lanes 1-16) was incubated with 0
nM (lanes 1 9), 0.5 nM (lanes 2 10), 1.0 nM
(lanes 3 11), 2.0 nM (lanes 4 12), 4.0 nM (
lanes 5 13) 8.0 nM (lanes 6 14) 16.0 nM
(lanes 7 15) and 32.0 nM (lanes 8 16) of ERß
and 400 nM HMGB-1 (lanes 9-16). In each set of
lanes, the amount of complex increased while the
amount of free DNA decreased with increasing
concentrations of ERß. The increase in complex,
and consequently the decrease in free DNA, occurs
at lower concentration of ERß in the presence of
HMGB-1, supporting the idea that HMGB-1 enhances
the binding affinity of ERß to cERE.
22

• Figure 11 Equilibrium binding profile of ERß
  binding to cERE in the presence and absence of
  HMGB1 The percent ERß complex bound in the
  presence (red circular dots) or absence (black
  square dots) of HMGB-1 obtained from EMSA
  profiles are plotted as a function of ERß. From
  this graph, the KD of ERß binding to cERE in the
  presence and absence of HMGB-1 can be estimated
  at 2.1 and 7.5 respectively.

23
ERaTable IV KD values for ERa/ERß to cERE in
the presence and absence of HMGB-1
24
Conclusions

• From the data presented above, it can be
  concluded that the isolation and purification of
  HMGB-1 was performed successfully with a level of
  purity and known concentration that made it
  functional for EMSA studies. Furthermore, the
  data collected from EMSA studies shows that
  HMGB-1 did enhance the binding of ERa/ERß to its
  consensus response element in a 33 base pair
  piece of double stranded DNA.