*3.2. Annellated Pyrroles*

Annellated pyrroles are prevalent in nature. For example, many well-known biologically active alkaloid families, including the lamellarins and indolizidins, as well as many stemona alkaloids, feature annellated pyrrole moieties [283–285].

Between 2010 and 2012, the highly halogenated 5- and 8-ring annellated pyrroles **326**–**328** were isolated from marine bacteria (Figure 46). The *Pseudoalteromonas*-derived 2,3,5,7-tetrabromobenzofuro[3,2-*b*]pyrrole (**326**) displayed significant antimicrobial activity against methicillin-resistant *Staphylococcus aureus* (ATCC 43300, IC50 value of 1.93 μM ± 0.05 μM) [286].

**Figure 46.** Annellated halopyrroles **326**–**328** derived from marine bacteria.

The biologically active ( −)-chlorizidine A (**327**) was isolated from a marine *Streptomyces* sp. and exhibited noteworthy activity in a human colon cancer cytotoxicity bioassay with IC50 values of 3.2–4.9 μM (Figure 46) [287]. Interestingly, the alkaloid **327** completely lost its activity when both phenolic functionalities were methylated. The authors also mentioned that a series of derivatives lacking the key 5 *<sup>H</sup>*-pyrrolo[2,1-*a*]isoindol-5-one moiety led to inactivity, strongly suggesting its presence is indispensable for biological activity [287].

The structure of ( ±)-marinopyrrole F (**328**), isolated from a *Streptomyces* sp. in 2010, contains an unusual eight-membered ring (Figure 46) [250]. In contrast to its enantiopure metabolites, marinopyrroles C–E (**244**–**246**, see Figure 32), ( ±)-marinopyrrole F (**328**) was isolated in racemic form. With the help of chiral HPLC, the authors found out that enantioenriched **328** completely racemizes within 18 h, most probably caused by the fused ether ring lowering the barrier for atropisomerism. However, ( ±)-marinopyrrole F (**328**) was much less active against MRSA and HCT-116 (MIC90 value 3.1 μg/mL) compared to (−)-marinopyrrole C (**244**, MIC90 value 0.16 μg/mL) [250].

In 2018, 4-debromougibohlin (**329**) and 5-debromougibohlin (**330**) were isolated from a marine sponge *Dictyonella* sp. by the Berlinck group (Figure 47). Unfortunately, both compounds did not show any proteasome inhibitory activity in a respective assay [264].

**Figure 47.** Structures of 2,3-annellated marine pyrrole alkaloids **329**–**332**.

In 2019, a related halopyrrole alkaloid incorporating the carbamoylpyrrole-like core structure, 1- *N*-methylugibohlin (**331**), was isolated from the sea sponge *Agelas nemoechinata*, but did not show cytotoxic activity against K562, A549, HeLa, or HCT-116 cells in vitro (Figure 47) [262].

Longamide C (**332**), obtained from an organic extract of *Agelas nakamurai* in 2010, was isolated as a racemic mixture (Figure 47). However, ROESY correlations indicated a half chair conformation of the six-membered ring. Compound **332** did not show any promising antimicrobial or cytotoxic activity [234].

In 2017, the Lin group isolated stylisines A–F (**333**, **282**, **283**, **334**, **335**, **132**) from the marine sponge *Stylissa massa*, of which stylisine A, D, and E (**333**–**335**) feature an annellated bromopyrrole moiety (Figure 48) [133]. The absolute stereochemistry of compounds **334** and **335** was deduced from ECD experiments. However, no antibacterial activity was observed for all three compounds **333**–**335** [133]. One year later, 5-debromougibohlin (**330**, Figure 47) was isolated and erroneously presented as a "new" bromo alkaloid [264], since it has the same structure as stylisine A (**333**).

**Figure 48.** Stylisines A (**333**), D (**334**), and E (**335**) from the marine sponge *Stylissa massa*.

At this point, the stereoselective synthesis of (−)-stylisine D (**334**) reported by Petkovic and Savic in 2019 should be mentioned (Scheme 20) [288]. The synthesis commenced with an N-protection and propargylation followed by routine transformations to generate allene **337** in 55% yield over three steps. After installing Boc-L-proline (**338**) which furnished compound **339** possessing the right configuration, compound **340** was obtained over four steps under transfer of chirality. After bromination and hydrolysis, a final oxidation step delivered (−)-longamide B (**341**), another bromopyrrole isolated from the sponge *Stylissa massa*. (−)-Stylisine D (**334**) was obtained by amidation of the carboxylic group of (−)-longamide B (**341**).

**Scheme 20.** Synthesis of stylisine D (**334**) and intermediate longamide B (**341**) via a metal-catalyzed cyclisation of allene **339** in a stereoselective manner.

In 2016, the family of longamides was extended by the isolation of longamides D–F (**342**–**344**) from a marine sponge *Agelas* sp. (Figure 49) [66]. Compounds **342**–**344** were isolated as racemic mixtures which were separated into pure enantiomers. The absolute stereochemistry of **342**–**344** was then determined by chiral HPLC and ECD spectroscopy. In the *Caenorhabditis elegans* candidiasis model, metabolites (+)-**342**, (−)-**343** and (+)-**344** exhibited significant antifungal activity with survival rates around 50%, whereas the corresponding enantiomers (−)-**342**, (+)-**343** and (−)-**344** did not show any activity, strongly suggesting the absolute configuration at C-9 to have an appreciable effect [66].

**Figure 49.** Longamides D–F (**342**–**344**) from the South China Sea sponge *Agelas* sp.

In 2014, several structurally unique annellated halopyrroles **345**–**348** were isolated from the Patagonian bryozoan *Aspidostoma giganteum* by Palermo and co-workers (Figure 50) [239]. The absolute configurations of bromotryptophan-derived aspidostomides D (**345**) and E (**346**) were determined by a modification of Mosher's method in combination with NOE correlations. While the elimination product of **345** and **346**, aspidostomide F (**347**), the N–N-linked dimeric aspidazide A (**348**) and compound **345** only exhibited moderate to weak cytotoxic activity against the 786-O human renal carcinoma cell line (IC50 values between 27.0 μM and >100 μM), aspidostomide E (**346**) proved active with an IC50 value of 7.8 μM [239].

**Figure 50.** New aspidostomides D–F (**345**–**347**) and aspidazide A (**348**) from the patagonian bryozoan *Aspidostoma giganteum*.

In 2017, a new family of annellated halopyrroles, the callyspongisines, were isolated from the Great Australian Bight marine sponge *Callyspongia* sp. (CMB-01152) (Figure 51) [289]. In callyspongisines A (**349**), a very rare imino-oxazoline core is spirocyclic to a seven-membered ring contiguous to a pyrrole unit. Due to insufficient quantities of **349**–**352**, the stereochemistry could not be determined and the authors also mentioned that callyspongisines B–D (**350**–**352**) could be storage and handling artifacts of **349** instead of being of natural origin [289]. The potent kinase inhibitory activity observed in *Callyspongia* sp. was attributed to hymenialdisine, while compounds **349**–**352** did not show any cytotoxic activity against a range of prokaryotic, eukaryotic, and mammalian cell lines [289].

**Figure 51.** Callyspongisines A–D (**349**–**352**) and pyrrololactam **353** of which only compound **349** is speculated to be of natural origin.

A related pyrrolactam alkaloid, axinelline B (**353**), was isolated from the *n*-BuOH extract of a marine sponge of the genus *Axinella* in 2017 (Figure 51). Unfortunately, the authors did not give any information about the stereochemistry or biological activity of compound **353 [131]**.

Annellated Pyrrole (Amino)-Imidazole Alkaloids

Several contiguous tetracyclic brominated pyrrole-imidazole alkaloids **354**–**356** were isolated or synthesized between 2016 and 2019.

In 5-bromophakelline (**354**), isolated from an Indonesian marine sponge of the genus *Agelas*, the relative and absolute configuration was deduced with the help of NOESY correlations and X-ray crystallography (Figure 52). However, no antimicrobial activity against *Mycobacterium smegmatis* (NBRC 3207), a model organism for tuberculosis was observed [290].

**Figure 52.** Brominated pyrrole-imidazole alkaloids **354**–**356** bearing guanidine units.

Compound **355** was isolated from the sponge *Agelas nemoechinata* in 2019 (Figure 52). The relative and absolute configuration of 9-*N*-methylcylindradine A (**355**) was determined by NOESY correlations and by the comparison of its optical rotation with the known (+)-cylindradine A. Unfortunately, no cytotoxic activity against K562 and L-02 cell lines could be observed [262].

At this point, we would also like to mention the first total synthesis of (+)-cylindradine B (**356**) (Scheme 21) [291], which was isolated from the marine sponge *Axinella cylindratus* back in 2008 [292].

**Scheme 21.** First total synthesis of (+)-cylindradine B (**356**) via key Pictet–Spengler reaction.

The authors commenced their synthesis with prolinol derivative **357** which was transformed with pyrrole **358** into the Pictet–Spengler precursor **359** over several steps. The Pictet–Spengler reaction then selectively gave compound **360** under addition of (±)- 1,1--binaphthyl-2,2--diyl hydrogen phosphate. In the next steps, the guanidine group was attached via an isothiourea intermediate **361**, which reacted with NH3/MeOH furnishing compound **362**. After changing the protective groups, the Boc-protected pyrrole **363** was brominated by using bromine and a final deprotection by applying TFA furnished (+)- cylindradine B (**356**) in 14% yield over four steps (Scheme 21) [291].

In 2010, a compound very similar to **354**, dibromohydroxyphakellin (**364**), was isolated from *Agelas linnaei* and represents the first described 12-OH analog of the phakellin family (Figure 53) [234]. By comparison of its optical rotation data with those of related compounds, it was assumed that dibromohydroxyphakellin (**364**) was isolated as a scalemic mixture. No cytotoxicity was observed against the murine L1578Y mouse lymphoma cell line [234].

**Figure 53.** Structurally complex annellated bromopyrroles **364**–**369** isolated from *Dyctionella* sp. or *Agelas* sp.

In 5-bromopalau'amine (**365**), isolated from *Dictyonella* sp. (marine sponge), the relative configuration of the eight stereogenic centers was determined by ROESY correlations [264] and was in accordance with the data reported for the revised structure of palau'amine (Figure 53) [293]. Compound **365** displayed proteasome inhibition activity with an IC50 value of 9.2 μM ± 3.2 μM, whereas the debrominated analog, palau'amine, was fourfold more active. Due to these data, the authors mentioned that both, bromination and the position of the bromine substituent in the pyrrole moiety seem to significantly influence the ability to inhibit the 20S yeas<sup>t</sup> proteasome [264].

In 2019, a new class of annellated bromopyrroles, the agesamines A (**366**) and B (**367**), were isolated as an inseparable epimeric mixture from an Indonesian sponge of the genus *Agelas* (Figure 53). The absolute configuration of both compounds **366** and **367** was elucidated by ECD measurements [294].

The related agelastatins E (**368**) and F (**369**) were isolated from the marine sponge *Agelas dendromorpha* in 2010 (Figure 53) [295]. The relative configuration of both compounds **368** and **369** was determined by NOESY correlations and by comparison to the known congener agelastatin A. As agelastatin A is a highly cytotoxic compound, agelastatins E (**368**) and F (**369**) were screened for cytotoxicity against the human KB cell line. Unfortunately, both compounds **368** and **369** lacked significant activity [295].

Concerning the agelastatin family, the total synthesis of agelastatins A–F (**375**–**378**, **368**, **369**), published by the Movassaghi group in 2010, should be mentioned (Scheme 22) [296]. The synthesis commenced with the known pyrrole **370**, which was converted into the annellated pyrrole **371** in 62% yield over four steps. After the addition of a stannylmethylurea in the presence of Liebeskind's CuTC reagen<sup>t</sup> and treatment with methanolic HCl, (+)-*O*-Mepre-agelastatin A (**372**) was obtained. Subsequent heating in aqueous methanesulfonic acid then furnished the natural product agelastatin A (**375**) in 49% yield as well as a side product (**374**). Bromination or OH-methylation of agelastatin A (**375**) gave agelastatin B (**376**) or E (**368**), respectively. Moreover, (−)-*O*-Me-di-*epi*-agelastatin A (**374**) could be further converted to agelastatin C (**377**) by an elimination/epoxidation/aqueous epimerization sequence. By reacting the former intermediate **371** with a stannylurea, agelastatins D (**378**) and F (**369**) could be synthesized in a similar way (Scheme 22) [296].

**Scheme 22.** Enantioselective synthesis of all known (−)-agelastatins, including the first total synthesis of agelastatins C–F (**377**, **378**, **368**, **369**).

In 2020, a new member of the agesamine family, agesamine C (**379**), could be isolated from the sea sponge *Agelas oroides* collected off the Tel Aviv coast (Figure 54). The relative and absolute configuration of the bicyclic moiety in **379** was deduced by comparison of its *J*-values with those of agesamines A (**366**) and B (**367**) [237].

**Figure 54.** Structurally diverse bromopyrrole alkaloids **379**–**383** isolated from *Agelas oroides*.

Monobromoagelaspongin (**380**) was first isolated from the sponge *Agelas oroides* as a racemic mixture in 2017 and no information was given on the relative configuration or its biological activities [297]. However, in 2020, the relative and absolute configuration could be determined alongside the isolation of further bromopyrroles (Figure 54) [237].

The same sponge also delivered the agelaspongin analogs **381** and **382**, the relative and absolute configurations of which were either determined by NOESY data combined with ECD spectroscopy or by comparison of its chiroptical properties with those of model compounds (Figure 54) [237]. The sponge also was the source of a new compound, named dioroidamide A (**383**). Compound **383** presents a negative specific rotation value which is also the case for many other structurally related marine alkaloids, and based on their shared biosynthesis, the authors assumed that **383** should possess the same absolute configuration as depicted in Figure 54 [237]. With the isolated natural products **379**–**383** itself, no biological tests were performed. However, as the antimicrobial and antibacterial activity of the sponge extract was attributed to other natural products contained, compounds **379**–**383** have not been found to show any promising activities so far [237].

In 2014, two structurally unique dimeric bromopyrroles, named agelamadins A (**384**) and B (**385**), were isolated from a sponge of the genus *Agelas* by the Kobayashi group [298]. Both compounds **384** and **385** were isolated as racemic mixtures, with their relative configurations determined by ROESY correlations (Figure 55). Agelamadins A (**384**) and B (**385**) showed antimicrobial activity against several Gram-positive species with IC50 values ranging between 4 μM and 16 μM. However, no cytotoxicity was observed against human murine lymphoma L1210 cells and human epidermoid carcinoma cells in vitro [298].

**Figure 55.** Annellated bromopyrroles **384**–**390** from different marine sponges.

Two new bromopyrroles **368a** and **368b**, annellated by a seven-membered ring and structurally related to hymenialdisine, were isolated from the marine sponge *Cymbastela cantharella* in 2011 (Figure 55) [132]. The absolute structure of (+)-dihydrohymenialdisine (**368a**)

was unequivocally determined by X-ray crystallography, whereas the absolute configuration of ( −)-dihydrohymenialdisine (**368b**) could not be deduced. Since the corresponding lead structure, hymenialdisine, is active against the kinase PLK-1, both substances **368a** and **368b** were also tested for PLK-1 inhibition but did not show any activity. Apparently, the conjugation of hymenialdisine through the C-10/C-11 double bond (which is saturated in **368a** and **368b**) is indispensable for its strong activity on a wide range of cyclin-dependent kinases [132].

The structurally similar compounds **387**–**390** were isolated from a marine sponge of the genus *Stylissa* in 2012 (Figure 55) [299]. While 12- *N*-methylstevensine (**387**) displayed strong cytotoxic activity against L5178Y mouse lymphoma cells with an EC50 value of 3.5 μg/mL, 12- *N*-methyl-2-debromostevensine (**388**), 3-debromolatonduine B methyl ester (**389**), and 3-debromolatonduine A (**390**) only exhibited weak activity (no values given). These data sugges<sup>t</sup> that the presence or absence of bromine atoms significantly influences the antiproliferative activity [299].

At this stage, it should also be mentioned the recently published total synthesis of the related pyrroloazepinone-containing alkaloid 2-debromohymenin (**396**) (Scheme 23) [300]. First, the commercially available 4-iodoimidazole **305** was transformed into alkyne **391** by a Sonogashira reaction. Subsequent deprotection and reaction with pyrrolecarbonyl chloride **392** furnished compound **393**. An intramolecular gold-catalyzed alkyne hydroarylation then resulted in the formation of the core pyrroloazepinone moiety in **394**. Subsequent hydrogenation followed by the installation of an azide group generated azido derivative **395**. Bromination using NBS, removal of the sulfonyl urea, and final conversion of the azide to an amine group as well as removing the N-OMe group at the same time using Mo(CO)6, furnished 2-debromohymenin (**396**) [300].

**Scheme 23.** Total synthesis of 2-debromohymenin (**396**) via a key gold-catalyzed alkyne hydroarylation.
