Solution NMR backbone assignment reveals interaction-free tumbling of human lineage-specific Olduvai protein domains

Abstract

Olduvai protein domains, encoded primarily by NBPF genes, have been linked to both human brain evolution and cognitive diseases such as autism and schizophrenia. There are six primary domains that comprise the Olduvai family: three conserved domains (CON1-3) and three human lineage-specific domains (HLS1-3), which typically occur as a triplet (HLS1, HLS2 and HLS3).

Herein, we present the solution NMR assignment of the backbone chemical shifts of the separate HLS1, 2 and 3 domains of NBPF15. Our data suggest that there is no change in the structure of the separate domains when compared to the full-length triplet (HLS1-HLS2-HLS3). We also demonstrate that there is no direct interaction between the three domains.

Biological context

Olduvai protein domains (known formerly as DUF1220)(Sikela and van Roy 2017) are on average 65 amino acid segments that are encoded primarily by NBPF genes (Vandepoele et al. 2005; Popesco et al. 2006;). Sequences encoding the domain are highly amplified in humans and show the largest human-specific increase in copy number of any coding region in the genome (~300 total copies of which 165 are human-specific) (Dumas et al. 2012; O’Bleness et al. 2012). The domain family, encoded by approximately 20 human NBPF genes (Figure 1), has been classified into six subtypes based on sequence similarity: three subtypes that are more conserved across species, CON1-3, and three subtypes that are amplified specifically in humans, HLS1-3. The HLS subtypes, which show high copy number variation in humans, are typically organized as a three-domain block called the Olduvai triplet (O’Bleness et al. 2012; Sikela and van Roy 2017).

Figure 1.

Organization of Olduvai-encoding sequences in humans. Shown are different isoforms of NBPF genes with different copy numbers of the Olduvai triplet (HLS1-3).

Several lines of evidence point to the involvement of Olduvai in brain evolution. For example, Olduvai copy number shows a striking correlation with brain size, neuron number and measures of cognitive function (Dumas and Sikela 2009; Dumas et al. 2012; Keeney et al. 2014; Davis et al. 2015). In addition, it has been recently demonstrated that most Olduvai triplets are adjacent to, and co-regulated with three human-specific NOTCH2NL genes that have been implicated in promoting cortical neurogenesis and brain expansion (Fiddes et al. 2019). In contrast to these potentially beneficial effects, several studies have linked variation in Olduvai copy number with brain disorders including autism, schizophrenia, macrocephaly and microcephaly. For example, while an increase in CON1 copy number has been linearly associated with increased autism severity (Davis et al. 2014, 2015a, 2019), CON1 decrease has been linearly associated with schizophrenia severity (Searles Quick VB, Davis JM, Olincy A, Sikela JM 2015). The fact that Olduvai variation has been implicated in both beneficial and detrimental effects has led to the proposal that Olduvai domains may represent a cognitive genomic trade-off, and that it is which, where, how and when copies change that determines whether the effects are advantageous or harmful (Sikela and Searles Quick 2018).

Methods and experiments

Protein expression and purification

HLS1, HLS2, HLS3, and triplet HLS1-3 constructs derived from NBPF15 were cloned into the expression vector pET-21b(+) with a N-terminal maltose binding protein (MBP) tag coupled to a tobacco etch virus (TEV) protease recognition site and C-terminal 6xHistidine tag. The plasmids were transformed and expressed in Escherichia coli strain Rosetta™ 2(DE3)pLysS cells. Single colonies from overnight culture plates were resuspended in 3 mL of LB-Miller media and shaken at 37 °C for 8 hours. M9 media formulated with 1 g/L ¹⁵N-ammonium chloride and 1.5 g/L ¹³C-glucose was inoculated with the preculture (1000:1) and shaken overnight at 37 °C. M9 media with ¹⁵N-ammonium chloride and ¹³C-glucose was inoculated with the overnight culture (50:1) and shaken at 37 °C. The culture was induced with 0.5 mM isopropyl-1-thio-d-galactopyranoside once the A₆₀₀ reached 0.6 and shaken for 5 hours at 30 °C. Cells were harvested by centrifugation at 4 °C for 10 min at 4500 × g.

Cell pellets were resuspended in 50 mM Tris (pH 7.5), 300 mM NaCl, 2 % glycerol, 10 mM imidazole, and 10 mM β-mercaptoethanol. Cell disruption was performed by sonication (6 × 20 s) at 4 °C. The lysates were clarified by centrifugation at 20,000 × g for 30 min at 4 °C and loaded onto a pre-equilibrated column packed with Ni Sepharose excel resin (GE Healthcare). The column was washed with 5 column volumes (CV) of resuspension buffer and His-tagged protein was eluted with 50 mM Tris (pH 7.5), 300 mM NaCl, 3 % glycerol, and 400 mM imidazole. TEV protease was added to the pooled elution fractions and the resulting sample was dialyzed overnight at 4 °C against 10 mM Tris (pH 7.5), 100 mM NaCl, and 5 mM dithiothreitol (DTT). The protein sample was loaded onto a pre-equilibrated column packed with Source 15Q (GE Healthcare). Proteins were separated using the following gradient: 0 – 50 % B (0 – 5 CV), 50 – 75 % B (5 – 6 CV), 75 – 100 % B (6 – 6.5 CV), 100 % B (6.5 – 7.5 CV) with buffer A (20 mM Tris, pH 8.0) and buffer B (20 mM Tris, 1 M NaCl, pH 8.0). The relevant fractions were concentrated using a 3000 MWCO concentrator (Sartorius) and buffer exchanged into 50 mM potassium phosphate, 100 mM NaCl, and 2 mM DTT.

NMR spectroscopy

¹³C/¹⁵N-labeled NMR samples of NBPF15 HLS1, 2 and 3 domains were prepared in 50 mM potassium phosphate, 100 mM NaCl, and 2 mM DTT buffer with protein concentrations between 120 and 250 μM. Backbone assignment of the three domains was achieved using ¹⁵N-¹H HSQC, HNCACB, CBCA(co)NH, HNCO, HN(ca)CO experiments recorded on 900 and 600 MHz triple-resonance Varian cryoprobe spectrometers at 25 °C (Ikura et al. 1990)(Cavanagh 2007). The 3D spectra were acquired with a nonuniform sampling (NUS) scheme generated by NUS@HMS scheme generator (Hyberts et al. 2012b) employing 1024 complex data points in the direct dimension and 25-30% sampling of the original 96 and 80 points in the indirect ¹³C and ¹⁵N dimensions, respectively. For all experiments, the spectral widths were 14,045 Hz (¹H), 3,200 Hz (¹⁵N), 3,770 Hz (¹³C=0), and 15,835 Hz (¹³Cα/¹³Cβ), the number of scans 16 and 64 (or 128) for 3D experiments and for 2D HSQC, respectively, and the interscan delays 1.7 s. For NMR titration analysis, ¹⁵N-¹H HSQC spectra with 128 complex points in the indirect dimension and 128 scans were recorded at 25°C for 40 μM samples of ¹⁵N-labeled HLS2 alone and in the presence of 0.5 and 1 equivalents of unlabeled HLS1. The 3D NUS-spectra were constructed using the hmsIST software (Hyberts et al. 2012a), and the linearly acquired 2D spectra were subject to NUS zero-filling as an alternative to linear prediction. A solvent subtraction function was applied in the direct dimension. Further data processing and visualization were performed using NMRpipe/NMRDraw (Delaglio et al. 1995) and NMRFAM Sparky (Lee et al. 2015). Resonance assignment was performed using the CCPNmr analysis software (Vranken et al. 2005).

Assignment and data deposition

The ¹⁵N-¹H HSQC of the HSL1-3 triplet showed a peak dispersion characteristic of a disordered protein. In addition, the three HLS domains share at least than 67% sequence identity. Due to the resulting high sequence redundancy between all HLS domains in NBPF15 (Figure 2), we used a “divide-and-conquer” approach for chemical shift assignment. We divided the triplet into its three domains HLS1, 2 and 3, while keeping the adjacent linker sequences in all the constructs. Owing to the moderate dispersion of the resonances the assignments success was 90±5 % (Table 1) with HLS1 higher than HLS2 and 3. The three backbone assignments for HLS1, 2, and 3 have been deposited in the BMRB with accession codes 27569, 27533, and 27775, respectively.

Figure 2.

Top: Sequence alignment between the three assigned NBPF15-HLS domains (HLS1, 2, and 3). Residues corresponding to cloning artefacts and tags are highlighted with a black box, while relevant residues belonging to the protein are highlighted with a red box. Bottom: Assigned ¹H-¹⁵N HSQC spectra of HLS1, 2, and 3 (dark blue, violet, and cyan contours) overlaid with the spectrum of the full-length triplet construct (green contours).

Table 1:

Backbone assignment statistics of HLS1, 2, and 3.

HLS construct with numbering of relevant residues	Total number of relevant residues*	Total number of relevant non-proline residues	% backbone resonances assigned (number of backbone atoms assigned)
NBPF15-HLS1 338-412	75	71	94.3% (67 15N, 67 Cα, 64 Cβ, 67 CO)
NBPF15-HLS2 401-487	87	81	88.8% (72 15N, 72 Cα, 70 Cβ, 72 CO)
NBPF15-HLS3 478-552	75	68	86.7% (59 15N, 59 Cα, 57 Cβ, 58 CO)

^*All constructs have 19 extra non-relevant residues from cloning and the His-tag.

Chemical shift analysis

Overlays of the ¹⁵N-¹H HSQC spectrum of the full-length triplet with those of each individual domain show no change in the peak positions except for peaks corresponding to the linkers between the three domains (Figure 2A). This indicates absence of long-range interactions between the three HLS domains.

To confirm the lack of long interaction between the different HLS domains, we also ran NMR titration experiments of unlabeled HLS1 with ¹⁵N-labeled HLS2 (Figure 3). Our data showed no peak shifting or reduction of intensity in HLS2 upon addition of the HLS1 construct.

Figure 3.

HLS2 interaction with HLS1. Overlay of NMR ¹⁵N-¹H HSQC spectra of free ¹⁵N-labeled HLS2 (violet contours) and in 1:1 mixture with unlabeled HLS1 (blue contours). Note that the shifted peak at 125.5 ppm/8.15 ppm belongs to E486, which s not part of the domain, but of the truncated linker.

In a next step, we will use the resonance assignment to investigate if there is a direct interaction between the Olduvai triplet and the products of the three human-specific NOTCH2NL genes, which are co-regulated.