Sp1 and copper

This post investigates a claim in a review by Morrel and coworkers that excessive dietary fructose can decrease the expression of Ctr1 in the duodenum. [1] How could this happen? What is transcription factor for Ctr1? As it turns out, both Cu⁺ and Cu²⁺ can bind to the CTR1 transcription factor Sp1 and negatively regulate its activity [2,3] The featured image describes out Cu can disrupt the Zn²⁺ finger(s?) Sp1 uses to bind to the promoters of itself and CTR1. If you wish to skip the details of some brilliant experiments, click to go to the lay friendly summary.

Cu²⁺ prevents Sp1 promoter binding

In their introduction to their Cu²⁺ study Yan and coworkers discussed the role of cysteines and histidines in binding Zn²⁺ and other transition metals like cadmium and copper.  When metal ions are not bound to the zinc fingers of Sp1, Sp1 binds to the GC-rich boxes with the consensus sequence 5’-G/T-GGGCGG-G/A.  This GC box lies upstream of the gene for Sp1 itself, CTR1, and many other genes.  In the absence of copper, Sp1, the protein transcription factor, increases transcripts for more Sp1 and Ctr1 until copper balance is restored.   Figure 1 of this study used a nuclear extract of non small cell lung cancer cells that contained several isoforms of the SP1 protein.  This extract was mixed with 32P labeled double stranded DNA probes that have a negative charge.  The mixture was added to an agarose gel. An electrical current was applied such that the negatively charged small pieces of probe DNA migrated to the anode.   When bound to Sp1, they migrated slower.  Concentration ramps of  t 0, 1, 10, 50, 100, and 200 µM divalent metal ions were added to test the hypothesis that less probe would be present in the upper band that contains Sp1.  A Western blot was performed on the material in 1B to demonstrate that the upper band contains Sp1. 

Cu⁺ prevents Sp1 promoter binding

Technically speaking, there should be no free copper inside the cell.  It should be bound to some chaperone.  Cu should be in the Cu⁺ oxidation state. This is what is ground breaking in the Yuan 2017 study. Sp1, even when its Zn²⁺ fingers are metallated and ready to bind to CTR1 promoter, the chaperone Atox1 can donate a Cu+ and turn it off.  

1 Cu⁺ binds to all three Zn fingers

Yuan and coworkers kept Cu in the +1 oxidation state by use of the reducing agent TCEP.  The absorbance at 262 nm was taken as an indication of S- Cu⁺ charge haring transition. [3]  This particular experiment only used the second Zn²⁺ finger.  The second Zn²⁺ finger contains two tryptophans W560 and W571.  Tryptophan fluoresce at 340 nm when excited at 280 nm.  Fluorescence may be quenched when these aromatic residues are exposed to an aqueous environment, or as the authors speculated, by interactions of the thiolate anion with Cu⁺.  The authors also stated that the apo 2^nd Zn²⁺ finger tends to be unfolded.  The amino acid sequence of the 2nd Zn finger is shown in the featured image note that there’s a tryptophan flanked by two cysteines ( C yellow box).

Note that Cu binding data for just the 2^nd Zn finger saturate at 1:1 Cu per Sp1. For the full length Sp1, the binding saturates at a 3:1 ratio. This suggests that all three Zn fingers bind Cu.

2. Cu(I) can react with Zn-bound Sp1

 Figures 2A, s2 and s3 used a Cu⁺ chelating agent called bicinchoninic acid (BCA).  BCA has a strong absorbance maximum at 582nm when it binds to Cu⁺.  Panel 2A is not (just)  the absorbance spectrum of Sp1 with Cu⁺.  BCA was added to the material retained on an ultrafiltration membrane with a 3 kDa cutoff size.  Anything smaller is passed through the filter and larger is retained in a small amount of solution on the top of the filter.  As a negative control, the filtrate and retentate of Sp1 that had never seen Cu did not react with BCA to give the magenta color.  Table 1 shows the results.  Almost all of the 30 μM Cu⁺ is picked up by 10 μM Sp1.  About 2.μ.  There was 2.32 μM Zn²⁺ displaced from 10 M Sp1.  8.28 Zn²⁺ remained “in” the dialysis tube with the Sp1.  These data estimate the Zn²⁺ – Sp1 binding ratio to be 0.68 and 3.01 for Cu⁺ to Sp1. 

Panel 2B used a technique called mass spectrometry to measure the mass/charge ratio of the Sp1 2^nd Zn finger.  There is a semi clear reduction in mass when Cu⁺ displaces Zn²⁺  These two atoms are very close in mass.  To make matters more interesting histidines and cysteines have side chains that may be protonated, or not depending on the pH of the solution and whether they are binding transition metals.  Supplemental figure 2 shows use of 5 and 30% acetonitrile as a means of slowing down oxidation of Cu-BCA.  Supplemental figure 3 shows just the 2^nd Zn finger of Sp1 stealing Cu from CuBCA in the presence of 10% v/v acetonitrile.  In just buffer, the absorbance stays fairly constant.  As Zn-Sp1-zf2 is added, The absorbance of Cu-BCA decreases about one third of the original.  When there is no Zn, the decrease in Cu-BCA absorbance is essentially complete when there is twice as much 2^nd Zn finger as BCA bound Cu.  Just a note, UV /visible light absorption is often used to measure metal ions interacting with proteins.  These contributions are probably not that great in this particular system.

3. NMR experiments

The authors did not disclose details of how they labeled the expressed protein with ¹⁵N instead of the more abundant ¹⁴N isotope.  It is assumed that they supplemented the bacterial expression system with ¹⁵N ammonium chloride or something. The NMR technique has heteronuclear single quantum coherence (HSQC) that basically measures ¹H-¹⁵N pairs within the amino acids of the protein. Each dot on the graph represents a bonded H and N in only the ¹⁵N labeled protein.

Yuan and coworkers reported that the association constant  of Cu for K_Cu-Atox1 = 2.51x  10¹⁷protein.  Adding Apo Sp1 changes a number of N-H pairs.  Figure s4, going in the reverse direction changes little.  We are assuming that Cu-Atox1 is the ¹⁵N-labeled protein.  Figure s5 is somewhat of a control showing some massive, global changes when Zn²⁺ binds apo Sp1.  Note that these changes seem to be larger than when Cu⁺ binds apo Sp1.

4. Circular Dichroism, a window into protein secondary structure

Circular dichroism is a method involving the absorbance of circularly polarized light by structures within a protein.  It is used to measure the proportions of α-helices, β-sheets, and random structures.  The yellow to red helix in the featured image is considered an α-helix. Binding of a metal ion may be anticipated to make a protein more ordered. 

Yuan and coauthors considered that the addition of Cu⁺ and Zn²⁺ both resulted in a more ordered zf2.    The 2D HSQC NMR data in 4A and s6 also indicated not that much change between the two metal ions and the existence of a Cu finger.  Figure s6 is a mirror experiment using the entire Sp1 protein.  Zn-Sp1 + Cu⁺ seems to be somewhere in between the disordered apo Sp1 and Zn-Sp1 structures.  Cu⁺ disruption of GC box binding is what really matters in this story. 

5 Cu⁺ inhibits Sp1 from binding to GC boxes

Yuan and coworkers also used the gel mobility shift assay to demonstrate tha at about 3:1 moles Cu+ to 1 mole Sp1, the probe of the GC rich binding site on the CTR1 promoter ceases to bind Sp1 (top band, panel 5A. Panel s7A shows the opposite is true for Zn²⁺. These authors used a lot of NMR data

What about post translational changes in Sp1 other than metal binding?  Tan and Khachigian wrote an insightful review in this regard.[4] 

Some further investigation of the cited references revealed hexose kinase was also down regulated by glucose deprivation. [5] The PP1 inhibitor okadaic acid decreased the increased transcription of aldolase and pyruvate kinase in response to glucose. [5]  These authors saw what they interpreted as minor proteolysis of the dephosphorylated Sp1 in their Western blots. [5] 

Note that two protein kinase C phosphorylation sites flank Zn finger #2. Protein kinase C is activated by Ca²⁺.

Concluding remarks

We are little closer to understanding how fructose can lead to copper deficiency than we first began. Over dosing on Cu is still possible. It would seem that nature has given us a way to turn off import when we have enough. Sp1 even “knows” when our Cu chaperones are copper replete. [3] We are only scratching the surface of possible Sp1 post transnational modifications that might impact Sp1 binding to the CTR1 promoter. It is becoming apparent that the cure for copper deficiency may not be as simple as eating more copper.

• S-glutathionylation is a postranslational modification that occurs during oxidative stress whereby glutathion forms S-S bonds with protein thiols. Recall that the SH group of Zin finger cysteines. Glutathionylation may be reversed with NADH dependent enzymes thioredoxin and glutaredoxin.
• Advanced glycation end products. AGE are formed when -NH2 on side chains of amino acids such as lysine react with the aldehyde groups of sugars, fructose more than glucose. Note that there are a few lysines (K) in the sequence of the 2^nd Zn Finger of Sp1.
• N-acetylation of lysines may occur on K2 and K703 of human Sp1 according to UniProt.org. K703 is within the 3^rd Zn finger. We at CopperOne think that a copper replete, functioning mitochondria is needed for generation of NAD⁺ which regulates the deacetylation enzyme Sirt1.

Lay_Summary

Sp1 binds upstream of the protein coding parts of its own gene and the gene for the Ctr1 transporter. It contains three Zinc fingers. Cu+ and Cu2+ can displace the Zn and stop Sp1 from making more Sp1 and Ctr1. Sp1 can have phosphates attached to it by enzymes called kinases. Phosphatases take phosphates off. Phosphates some how or another may increase transcription of down stream genes. Speculation is that fructose may activate a phosphatase.. We’ve still no clue as to how fructose causes copper deficiency… just some educated hypotheses.

For CopperOne critics who don’t like us using Dr Brewer as a reference, we are remaining fast in our conviction that Cu²⁺ is bad. A new colleague shares this conviction. If Cu²⁺ is absorbed via the divalent metal ion transporter, the right way to absorb proper Cu⁺ gets shut down.

References

1. Morrell, A., Tallino, S., Yu, L., & Burkhead, J. L. (2017). The role of insufficient copper in lipid synthesis and fatty-liver disease. IUBMB life, 69(4), 263–270. PMC free article
2. Yuan S, Chen S, Xi Z, Liu Y. Copper-finger protein of Sp1: the molecular basis of copper sensing. Metallomics. 2017 Aug 16;9(8):1169-1175. PMC free article
3. Yan, D., Aiba, I., Chen, H. H., & Kuo, M. T. (2016). Effects of Cu(II) and cisplatin on the stability of Specific protein 1 (Sp1)-DNA binding: Insights into the regulation of copper homeostasis and platinum drug transport. Journal of inorganic biochemistry, 161, 37–39. PMC free article
4. Tan, N. Y., & Khachigian, L. M. (2009). Sp1 phosphorylation and its regulation of gene transcription. Molecular and cellular biology, 29(10), 2483–2488. PMC free article
5. Schäfer D, Hamm-Künzelmann B, Brand K. (1997) Glucose regulates the promoter activity of aldolase A and pyruvate kinase M2 via dephosphorylation of Sp1. FEBS Lett. 1997 Nov 17;417(3):325-8. free article