HomeLatest ThreadsGreatest ThreadsForums & GroupsMy SubscriptionsMy Posts
DU Home » Latest Threads » Forums & Groups » Topics » Science » Science (Group) » Structural basis of nucle...

Wed Sep 18, 2019, 08:06 PM

Structural basis of nucleosome recognition and modification by MLL methyltransferases

The paper I'll discuss in this post has the title I've used for the post itself: Structural basis of nucleosome recognition and modification by MLL methyltransferases (Jing Huang et al Nature 573, 445–449 (2019))

Last night I had the pleasure of attending a lecture by Dr. Benjamin Garcia who is a world expert in the structure of chromatin, specifically, the epigenetic implications of particular post translational modifications of histones. Histones are the proteins that wrap DNA, choreographing the way they and the genes that constitute them turn on and off, to simplify the matter somewhat. The histones - there are four of them - are extremely basic proteins, inasmuch as they are rich in arginine and lysine, and it is the chemistry of the latter amino acid, lysine, that is a controlling factor in how histones behave and function.

(As aside, the evening was an embarrassment of riches, Dr. Garcia's lecture was followed by one by Dr. Vicki Wysocki - in the picture of her lab group she is in the back row on the far left, partially obscured by one of her students. Dr. Wysocki is a world leader in the use of mass spectrometry to study protein complex structures, that is how proteins interact with one another to conduct the business of metabolism. She showed work where she defined the structure of an hexameric protein by mass spectrometry that was later confirmed by cryo EM imagining (the imaging was problematic so her group took a shot at it by mass spec) - no small trick.)

Anyway, about histones: As they control the operation of DNA, they are obviously involved in many processes involving cell division, both normal cell division and abnormal cell division, notably cancer. It is known that specific residues, in particular the ε-amino group in lysine, in the protein sequences of the histones are modified generally (but not always) in one of two ways, by acetylation or methylation. A bit of nomenclature: The term "H3K4" refers to histone 3, "H3" having a lysine residue (K in peptide language) in the 4 position in the amino acid sequence. The term "H2BK120" refers to histone 2B's lysine in the 120th residue of the sequence.

The histones in turn can also choreograph or allow the chemical modification of DNA itself - DNA can be methylated - this is the area of "epigenetics" which controls many areas of cell function and behavior, including, it seems, aging.

I've played in this space professionally. It's fascinating.

Before producing excerpts of the texts, it is probably useful to produce a graphic from the paper showing what this chromatin complex of histones and DNA looks like:

The caption:

a, Schematic of the domain organizations of the human MLL1 catalytic module. The colour scheme is the same as that of the MLL1–ubNCP structural model shown in c. DBM, DPY30 binding motif; PHD-WH, plant homeodomain and winged-helix domain; WIN-AS, WDR5-interacting motif and activation segment. b, Atomic model of human MLL1 catalytic module, shown from the cryo-EM structure of human MLL1–ubNCP complex. c, d, Cryo-EM density map (c) and atomic model (d) of human MLL1–ubNCP complex, shown from two orthogonal views. The cryo-EM map is segmented and coloured according to the components of the MLL1–ubNCP complex. Ub, ubiquitin.

The abstract and introduction describes what some of these abbreviations mean:

Methyltransferases of the mixed-lineage leukaemia (MLL) family—which include MLL1, MLL2, MLL3, MLL4, SET1A and SET1B—implement methylation of histone H3 on lysine 4 (H3K4), and have critical and distinct roles in the regulation of transcription in haematopoiesis, adipogenesis and development1,2,3,4,5,6. The C-terminal catalytic SET (Su(var.)3-9, enhancer of zeste and trithorax) domains of MLL proteins are associated with a common set of regulatory factors (WDR5, RBBP5, ASH2L and DPY30) to achieve specific activities7,8,9. Current knowledge of the regulation of MLL activity is limited to the catalysis of histone H3 peptides, and how H3K4 methyl marks are deposited on nucleosomes is poorly understood. H3K4 methylation is stimulated by mono-ubiquitination of histone H2B on lysine 120 (H2BK120ub1), a prevalent histone H2B mark that disrupts chromatin compaction and favours open chromatin structures, but the underlying mechanism remains unknown10,11,12. Here we report cryo-electron microscopy structures of human MLL1 and MLL3 catalytic modules associated with nucleosome core particles that contain H2BK120ub1 or unmodified H2BK120. ...

This is a paper about the epigenetics of leukemia.

...The human MLL1 catalytic module, which is composed of full-length WDR5, RBBP5, ASH2L and DPY30 (WRAD) proteins and MLL1 (residues 3754–3969), formed complexes with unmodified nucleosome core particles (NCPs) or NCPs containing mono-ubiquitinated H2BK120 (hereafter, ubNCPs) in electrophoretic mobility shift assays (Fig. 1a, Extended Data Fig. 1a, b). Cryo-electron microscopy (cryo-EM) single-particle analysis yielded a global density map of the MLL1–ubNCP complex at an overall resolution of 3.2 Ĺ (Extended Data Figs. 1c, 2, Extended Data Table 1). Another cryo-EM dataset of the MLL1–ubNCP complex was processed to an overall resolution of 4.0 Ĺ, and revealed electron microscopy density immediately N-terminal to SPRY domain of the ASH2L region (known as the pre-SPRY domain) as well as the conformational dynamics of ubiquitin within the complex (Extended Data Fig. 3a). We generated an atomic model of the MLL1–ubNCP complex by docking available high-resolution structures of human MLL1 subunits, the nucleosome and ubiquitin9,13,14,15,16 into the electron microscopy map, followed by manual building (Fig. 1b–d).

The "complexes" here are what Dr. Wysocki's group studies, not by imaging (Cryo-EM) but be interference from mass spectrometry data. Very cool.

Anyway, some more text.

The WD40 domain of RBBP5 (hereafter, RBBP5WD40) is sandwiched between the H2BK120-conjugated ubiquitin and core histones, and confers both nucleosome and ubiquitin recognition (Fig. 2a). The RBBP5WD40 is wedged into the H2B–H4 cleft, which represents a histone surface that is exploited by other nucleosome-binding proteins (such as Sir3 and 53BP114,20) (Fig. 2a, b). Two loops, which connect the WD40 propeller blades 5, 6 and 7, mediate direct interactions with nucleosome (Fig. 2b). One of these loops is accommodated between the C-terminal helix of H2B and nucleosome DNA adjacent to this helix, with the hydrophobic residues Leu248 and Val249 facing the C-terminal helix of H2B and the positively charged Arg251 pointing to the phosphate backbone of the DNA (Fig. 2b, Extended Data Fig. 4a). The other loop adheres to residues Lys79 and Thr80 of histone H3, which further stabilizes the association of RBBP5WD40 with the H2B–H4 cleft (Fig. 2b, Extended Data Fig. 4b). Removal of RBBP5WD40 severely impaired the methyltransferase activities of MLL1 towards nucleosomal H3K4 (Fig. 2d), which emphasizes that the recognition of a specific histone surface through RBBP5WD40 is required for efficient H3K4 methylation on the nucleosome.

RBBP5 is a retinoblastoma binding protein that is important in cell division; it is a tumor suppressor gene. (One of the interesting points of Dr. Garcia's talk last night concerned the interaction of histone tails with protein sequences that seem to be involved in immune function - the immunology of cancer is another interesting topic.)

I recognize that this is all very esoteric, and I produce it because of the high the lectures gave me, so let me limit the rest of this post to some pretty pictures of chromatin histone/DNA complexes with captions.

The caption:

a, Overview of the RBBP5–ubNCP interactions, showing that RBBP5WD40 is sandwiched between the histone H2B–H4 cleft and H2BK120ub1. b, Detailed view of the recognition interface between RBBP5WD40 and the histone H2B–H4 cleft. H2BαC, C-terminal α-helix of H2B. c, Interface between the α-helix-containing loop of RBBP5WD40 and H2BK120ub1 in the major RBBP5–ubiquitin binding mode, shown in front and back views. d, End-point histone methyltransferase (HMT) assays of equal amount of wild-type (WT) complexes and mutant MLL1 complexes (deletion of RBBP5WD40). Each assay was repeated at least three times with similar results. n = 3 independent experiments. Data are mean ± s.d. The input of the HMT reactions is shown in Extended Data Fig. 4f. e, Overview of the interactions of MLL1SET–ASH2L with the nucleosome, shown in front and back views. The ASH2L pre-SPRY domain (ASH2Lpre-SPRY) is shown with its electron microscopy density map. ASH2LSPRY, SPRY domain of ASH2L; C-ter, C terminus; RBBP5AS-ABM, activation segment and ASH2L-binding motif of RBBP5. f, Interaction between the N-terminal motif of MLL1SET (SET-N) and the C-terminal helical region of histone H2A. α2 and α3 denote α-helices 2 and 3 of histone H2A, respectively. g, HMT assays of an equal amount of MLL1 complexes with wild-type or truncated ASH2L. Each assay was repeated at least three times with similar results. n = 3 independent experiments. Data are mean ± s.d. The input of the HMT reactions is shown in Extended Data Fig. 5c. h, Electrophoretic mobility shift assays of MLL1 complexes bearing ASH2L truncations with NCPs at molar ratios of 1:1, 2:1 and 4:1. Each assay was repeated at least three times with similar results. The input of the assays is shown in Extended Data Fig. 5d. Numbers on the left represent the number of base pairs (converted from molecular mass). i, Alanine-scanning mutagenesis of the ASH2L pre-SPRY domain to identify residues that are critical for the activity of MLL1. Each assay was repeated at least three times with similar results. n = 3 independent experiments. Data are mean ± s.d. The input of the HMT reactions is shown in Extended Data Fig. 5e.

The caption:

a, Cryo-EM density maps and atomic models of the two binding modes of the human MLL1–NCP complex, in the dyad view of the nucleosome. b, Superposition of the two MLL1–NCP structures from a, showing the rotation of the MLL1 complex on the nucleosome surface between the binding modes. c, Detailed view of the recognition interface between RBBP5WD40 and SHL2-adjacent regions of the nucleosome from MLL1–NCP binding mode 2, shown with electron microscopy densities of key residues of the interface. d, Michaelis–Menten kinetic analysis of activities of human MLL1 on NCP and ubNCP. Data are mean ± s.d. n = 3 independent experiments. The kcat and Km of human MLL1 on ubNCP and NCP are shown with their standard errors.


The caption:

a, Cryo-EM density map and atomic model of human MLL3–ubNCP complex, shown from two orthogonal views. The colour scheme is the same as that of the MLL1–ubNCP complex, except that MLL3 is coloured in teal. b, Structural comparison of human MLL1 (coloured in orange) and MLL3 (coloured in teal) catalytic modules, based on structural superposition of MLL1SET and MLL3SET. c, Overall structural organization of the interface between WDR5, MLL1 and RBBP5 subunits. The WIN and activation segment motifs of MLL1 are shown with electron microscopy density maps. d, Detailed view of the interface between WDR5, MLL1 and RBBP5, highlighting a cluster of hydrophobic residues of MLL1AS and RBBP5AS that flank the insertion motif (labelled SET-I) and the SAH-binding pocket of MLL1SET. e, HMT assays performed with wild-type MLL1 complex and mutant MLL1 complex with deletions of both MLL1AS and RBBP5AS. Each assay was repeated at least three times with similar results. n = 3 independent experiments. Data are mean ± s.d. The input of the HMT reactions is shown in Extended Data Fig. 8c. f, Overview of the interface between WDR5, MLL3 and RBBP5 subunits. g, HMT assays of equal amount of MLL1 complex (MLL1–WRAD), MLL3 complexes with or without WDR5 (MLL3–WRAD and MLL3–RAD, respectively) and an MLL3 chimaera composed of the MLL1WIN-AS motif and the MLL3SET domain (MLL1AS–MLL3–WRAD) on the substrate of the nucleosome. Each assay was repeated at least three times with similar results. n = 3 independent experiments. Data are mean ± s.d. The input of the HMT reactions is shown in Extended Data Fig. 9b. h, A working model shows the different structural organizations at the interface between WDR5, MLL1SET (or MLL3SET) and RBBP5 that contribute to activity specificity in MLL1 (or MLL3) complexes. SAM, S-adenosyl-L-methionine.

From the paper's conclusion:

In summary, cryo-EM structures of human MLL1–ubNCP, MLL1–NCP and MLL3–ubNCP complexes provide a structural framework for nucleosome recognition and activity specificity of methyltransferases of the MLL family. The association between MLL proteins and nucleosomes ensures the proper deposition of H3K4-methyl marks in euchromatin regions, in which the open chromatin structures favour the access of MLL enzymes to the nucleosome surface. This association also allows the regulation of the activities of MLL enzymes by pre-existing histone marks or chromatin-binding factors, as exemplified in the trans-histone crosstalk between H2BK120ub1 and H3K4 methylation. These findings—together with recently reported cryo-EM structures of other histone modifiers (DOT1L and PRC2) in complex with nucleosomes25,26,27,28,29,30—highlight the importance of structural characterizations of histone-tail modifications in the context of nucleosomes and chromatin, which will shed light on the molecular mechanism(s) that underlie the complex and integrated regulation of histone modifications by chromatin structures and other epigenetic signals. Future studies are required to further investigate how an intact MLL complex works in the context of hierarchical chromatin structures.

Well, to each his own; I guess you had to be there.

Even cancer can have its own magnificence.

Evenings like last evening make my life worth living, such beautiful science albeit in a sadly dying world.

I hope your work week has had as many moments of transcendent joy as mine has had.

2 replies, 696 views

Reply to this thread

Back to top Alert abuse

Always highlight: 10 newest replies | Replies posted after I mark a forum
Replies to this discussion thread
Arrow 2 replies Author Time Post
Reply Structural basis of nucleosome recognition and modification by MLL methyltransferases (Original post)
NNadir Sep 2019 OP
trotsky Sep 2019 #1
NNadir Sep 2019 #2

Response to NNadir (Original post)

Thu Sep 19, 2019, 02:13 PM

1. I understand some of those words!

Thanks for the detailed post though. Science is amazing, and is really our only hope to save ourselves if we can. Figuring out how diseases work lets us figure out how to stop them from working. So glad there are smart people working on this.

Reply to this post

Back to top Alert abuse Link here Permalink

Response to trotsky (Reply #1)

Fri Sep 20, 2019, 06:30 PM

2. Thank you for reading. I make it habit to try to read as many things with title words I don't...

...understand as often as I have time to do so.

It's a very good habit, and can lead to awesome perspectives. At a minimum, it broadens one's horizons.

I'm still high from those lectures early in the week. Yes, there are very, very, very smart people, and the two I mentioned are among the smartest in their field, advanced biological mass spectrometry.

You are, in my opinion, serving yourself well simply by looking.

I will tell you that some of what was in this paper was over my head initially, but digging into it, I learned a lot.

Thanks again.

Reply to this post

Back to top Alert abuse Link here Permalink

Reply to this thread